## _wp15222

## Source details

**Canonical URL:** [_wp15222](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp15222.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp15222.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp15222.pdf.json)

---

### Introduction — overview and systemic relevance
- The U.S. Treasury market is one of the largest and most liquid financial markets and serves as the pricing benchmark for many financial instruments.
- Treasury securities: primary financing source for U.S. federal government deficits and key instrument for the Federal Reserve’s implementation of monetary policy.
- Uses: risk-free investment, hedging, margining, collateral; majority of official international reserves held in Treasuries.
- Average daily trading volumes:
  - Treasury market: US$500 billion.
  - U.S. equities: US$200 billion.
  - U.S. corporate bonds: US$25 billion.
- Systemic importance: disruptions in the Treasury market can affect auctions, trigger margin calls, and produce macro-financial ripple effects.
- Structural concern: “illusion of liquidity” — displayed liquidity can vanish quickly, causing higher price impact during low-liquidity episodes.

### Changes in market structure and implications for liquidity
- Electronic trading and HFT:
  - Most Treasury trading now occurs electronically via execution strategies and automated decision-making strategies.
  - Historical electronic adoption milestones:
    - 2001–2002: around 80 percent of benchmark Treasury bonds traded on electronic platforms.
    - End of 2004: more than 95 percent of inter-dealer trading of active issues on electronic platforms.
    - Treasury Borrowing Advisory Committee (2013): benchmark Treasury bonds and futures entirely traded on electronic platforms.
  - HFT market-share estimates:
    - HFT > 50 percent of trading activity on benchmark Treasury bonds and Treasury futures (estimates cited).
    - Cash market HFT activity: 40 to 50 percent of trading volumes.
    - Futures market HFT activity: >60 percent of volumes.
    - Reported share on BrokerTec among top 10 trading firms: dealers 14 percent; HFT firms 86 percent.
- Shift in traditional market makers:
  - From risk-warehousing (inventory-based) to risk-distribution (broker, pass-through).
  - Consequences: reduced shock-absorbing inventories; faster circulation of tradable risk; increased frequency of liquidity shortages and volatility spikes.
  - Drivers: technology, post-GFC business-model changes, regulatory and accounting changes (Leverage Ratio, Supplementary Leverage Ratio, Net Stable Funding Ratio, FASB 166 and 167, Basel III adoption).
  - Observed metrics:
    - Primary-dealer turnover of U.S. Treasuries fell to less than 5 percent (compared to 10 to 12 percent before the GFC).
    - Primary dealers reduced holdings of U.S. Treasuries since mid-2013; repo operations declined.
- Implications:
  - Displayed liquidity can be fragile; market orders can have outsized price impact during low-depth episodes, creating positive feedback loops that amplify shocks.

### Case study: October 15, 2014 “flash rally” — high-level findings
- Price and volume dynamics:
  - 10-year Treasury yield fell by 37 basis points (from 2.23 percent to 1.86 percent) intraday and rebounded, closing the day 6 basis points below the previous close.
  - Intraday moves of this magnitude had occurred only three times in the prior 25 years, all tied to significant policy announcements.
  - BrokerTec traded volumes surged to US$946 billion on October 15, more than 40 percent higher than the previous record of US$662 billion (May 2013).
- Liquidity evolution around retail sales release:
  - Around 25,000 contracts offered at top 10 prices for 10-year futures before release; liquidity vanished shortly before 8:30.
  - Liquidity recovered for about 20 minutes post-release but then declined rapidly from 8:50 through 9:30 to less than 20 percent of pre-announcement depth.
  - Post-event: market depth fell by 40 percent in Treasury bonds and by more than 50 percent in Treasury futures; partial recovery since then but remains below long-term averages.
- Tick-by-tick mechanics:
  - Price phases after 8:30 retail sales:
    - (i) 8:30–9:33: steady rise in futures prices (yields fall).
    - (ii) 9:33–9:40: sudden sharp increase in prices.
    - (iii) 9:40–10:00: sharp decline to more normal trading amid low liquidity.
  - Large aggressive buy trades at 9:33:45 included the day’s largest buy of 1,000 contracts, consuming about 10 percent of market depth at the 10 best prices.
  - Liquidity declined to less than 5,000 contracts at the 10 best prices during the episode; bid-ask spreads remained tight but quantities were very small.
  - Reversal driven by visible large sell limit orders (e.g., sell orders >1,000 30-year contracts represented 40 percent of 2,500 offered to buyers in futures; cash-market sell limits of US$8 million represented >70 percent of bonds offered to buyers) that provided resistance but were not executed.
- Safeguards:
  - CME safeguards (Stop Logic and Velocity Logic) did not trigger because prices moved continuously without discrete gaps; no single stop order execution met the Stop Logic criteria.

### Drivers and detailed contributors to the flash rally
- Repositioning by leveraged funds:
  - Background: Fed taper expectations earlier in 2014; leveraged funds had net short positions in interest rate futures to profit from rising rates.
  - In early October, global growth concerns and Ebola reports supported safe-haven flows; yields declined 15 to 19 basis points across 2- to 10-year maturities in the first two weeks of October.
  - On October 15, a FRB dealer survey found around two-thirds of respondents indicated hedge funds were net buyers of Treasuries between 8:30 and 10:00 AM.
  - Limitations: most hedge-fund repositioning occurred in September without large price impact; the flash dynamics and sharp reversal around 9:40 AM are not fully explained by hedge-fund buying alone.
- Withdrawal of liquidity providers:
  - Market depth was severely reduced; liquidity providers did not replenish consumed liquidity.
  - Depletion was larger for deeper order-book levels (6 to 10 best prices) than for levels close to best prices (1 to 5).
  - Indirect evidence indicates dealers reduced liquidity provision more than HFT firms; dealers may have reduced internalization (warehousing) and sent client orders to the market, amplifying buying pressure.
  - On control days average consecutive trades on same side: 20; on October 15 this fell to less than 10, consistent with reduced warehousing.
- High-Frequency Trading (HFT) metrics and interaction effects:
  - Order book updates rose from 10,000 updates per minute pre-event to a peak of 30,000 updates around 9:45 AM.
  - Trade sizes were unusually small: around half of the trades were for one contract — more than 20 percentage points higher than on October 2.
  - 10-Year Treasury futures contract notional value: US$100,000.
  - Reactive Trades (RTs) proxy:
    - RT estimates peaked at more than 16,000 contracts between 9:39 and 9:40, representing more than 25 percent of volume in that period.
    - During the event the share of trading by HFT firms increased to about 70 percent (versus 50 percent on control days per JSR findings).
  - Positive feedback loop: Low market depth → High number of trades → High price impact and order-book updates → High HFT activity → further depletion of depth.
- Authorities’ Joint Staff Report (JSR (2015)) findings:
  - Sharp decline in liquidity and elevated HFT activity were main contributors.
  - HFT firms reduced quantities offered; broker-dealers increased spreads and removed offers for brief periods.
  - Surge in message traffic increased latency; exchange IT systems faced operational risk.
  - Self-trades surged (up to 33 percent of volumes in the 5-year note during the event); JSR noted no legal findings on self-trading in the report.
  - JSR recommendations: review regulatory framework for Treasury securities; study potential registration requirements for HFT firms; reinforce best practices in the Treasury market.

### Policy measures and recommendations to strengthen market resiliency
- Three broad policy areas:
  - (i) provide incentives for liquidity provision;
  - (ii) improve market safeguards;
  - (iii) enhance regulatory framework and supervision of the Treasury market.
- Incentives for liquidity provision:
  - Consider introducing Designated Market Makers (DMMs) on Treasury cash and futures markets with contractual obligations to provide liquidity in exchange for privileges and possibly financial incentives.
  - Maker/taker fee structures on trading venues: require liquidity consumers to pay and offer liquidity providers a rebate — empirical effects mixed, further analysis needed.
  - Market-making requirements for Primary Dealers (PDs): possible quoting obligations, minimum bid-ask spread limits, minimum quoted volumes, or minimum presence during trading hours; include carefully structured waivers to avoid mass opt-outs.
  - International examples: many countries tie PD privileges to market-making obligations; U.S., Japan, and Germany currently have no secondary market obligations for PDs.
- Strengthen market safeguards:
  - Recalibrate existing safeguards (e.g., extend time window for Velocity Logic).
  - Include changes in market depth as an input to trigger safeguards in addition to price variation; example: the 75 percent decline in depth between 8:50 and 9:30 on October 15 could have been an explicit trigger.
  - Introduce market safeguards in Treasury cash market platforms (BrokerTec and eSpeed lack market safeguard mechanisms).
  - Harmonize safeguards across cash and futures markets while avoiding overly frequent halts that reduce price discovery.
- Regulatory and supervisory enhancements:
  - Expand regulatory perimeter: subject electronic trading platforms for Treasuries to specific regulation; supervise more participants to ensure a level playing field.
  - Address data gaps: create a consolidated tape of Treasury transactions to enhance market surveillance (no current reporting requirement comparable to equities/corporate bonds).
  - Supervision of HFT and proprietary trading firms:
    - Current supervision of proprietary firms is limited; those not registered as FCMs rely on FCMs’ risk controls.
    - SEC proposal (SEC (2015)) to require proprietary trading firms to become FINRA members.
    - JSR suggests studying registration requirements for HFT firms active in Treasuries.
- Concluding policy prescriptions:
  - Incentivize or mandate liquidity provision (e.g., PD obligations, DMMs).
  - Improve market safeguards including depth-based triggers and recalibrated velocity windows.
  - Strengthen regulatory and supervisory oversight, close loopholes, and fill data gaps to give authorities a complete view of the market.

### Appendix (estimation of Reactive Trades) — methodology and key model features
- Reactive Trade (RT) definition: a market order placed in reaction to a Best Quote Update (BQUD); used as proxy for algorithmic/HFT activity.
- Modeling approach:
  - Four mutually exclusive event types: RT, non-RT, non-simultaneous BQUD, simultaneous BQUD.
  - Intensity specifications:
    - Piecewise-constant λ_non-RT_i and λ_UD̅_i over intervals (t_{i−1}, t_i].
    - Reactive intensity separable: λ_RT(t, τ) = λ_RT_i g(t − τ) with g(h) = exp(−h^2 / 2σ^2), g(0)=1, g(h)→0 as h→+∞.
    - Probability q(t) of a BQUD conditional on a trade assumed piecewise-constant q_i.
  - Volume per trade:
    - V_RT and V_non-RT assumed Exponential(η_RT) and Exponential(η_non-RT), justified by fit for trades below 20 contracts (95 percent of trades).
- Estimation:
  - Parameters Θ = { σ, (λ_RT_i, λ_non-RT_i, λ_UD̅_i, q_i)_{i=1,...,n}, η_RT, η_non-RT } estimated by maximum likelihood using inter-event times and event-type indicators.
  - Practical setup: one-minute intervals for frequency parameters; smoothing via Kalman filter; robustness checks with 5-minute intervals.
- Simulation validation (one-minute period example):
  - True parameters: σ = 0.02, λ_RT = 300, λ_non-RT = 60, λ_UD̅ = 30, q = 0.3.
  - Over 100 simulated datasets mean estimated parameters:
    - σ: 0.02033 (SD 0.00091).
    - λ_RT: 296.0 (SD 8.8).
    - λ_non-RT: 59.29 (SD 8.78).
    - λ_UD̅: 29.91 (SD 0.70).
  - Conclusion: estimation performs well when data generated from model.
- Posterior classification and aggregate RT volume:
  - Prior RT probability: π(t) = λ_RT(t, τ_UD(t)) / (λ_RT(t, τ_UD(t)) + λ_non-RT(t)).
  - Trade-volume density mixture and posterior RT probability π_post(t) computed using exponential volume distributions.
  - Aggregate RT and non-RT volumes between times t and s:
    - A V_RT(t, s) = Σ_{h ∈ (t, s], trade at h} π_post(h) V(h).
    - A V_non-RT(t, s) = Σ_{h ∈ (t, s], trade at h} (1 − π_post(h)) V(h).
  - These aggregates provide a proxy for HFT-related activity shares.

*Source: IMF staff summary based on the provided content unit _wp15222.*

### Introduction ...........................................................................................................

### Introduction

### Overview and importance of the U.S. Treasury market
- The U.S. Treasury market is one of the largest and most liquid financial markets in the world and serves as the pricing benchmark for many financial instruments.
- Treasury securities are the primary source of financing for the U.S. federal government deficit and a key instrument for the Federal Reserve’s implementation of monetary policy.
- Treasury securities are widely used as a risk-free investment instrument and to satisfy hedging, margining, and collateral needs; the majority of official international reserves are kept in Treasury securities.
- Average daily trading volume in the Treasury market: US$500 billion.
- Comparative average daily trading volumes cited: U.S. equities US$200 billion; U.S. corporate bonds US$25 billion.

### Systemic relevance and risks
- Continuous functioning and smooth adjustment of the Treasury market are important to global asset markets and financial stability.
- Disruptions could have large effects with macro-financial implications, including effects on Treasury auctions and triggering margin calls with ripple effects through the financial system.
- Episodes of extreme intraday volatility can induce additional participants to trade in the same direction, amplifying shocks and producing destabilizing effects.

### The October 15, 2014 “flash rally” (high-level)
- On the morning of October 15, 2014, the yield on the 10-year Treasury bond fell by 37 basis points (from 2.23 percent to 1.86 percent) and rebounded quickly, closing the day only 6 basis points below the previous closing level.
- Intraday moves of this magnitude are highly unusual and, over the previous 25 years, had been observed only on three occasions, all driven by significant policy announcements.
- The event occurred in the absence of a major news event and was rapidly reversed; while it did not produce lasting systemic consequences, it may reflect structural vulnerabilities in market liquidity provision.

### Key characterization of current concerns
- Structural changes since the Global Financial Crisis have transformed liquidity provision: technological advances, the rise of new market makers, and changes in traditional market makers’ business models may create an “illusion of liquidity” — apparent liquidity that can vanish quickly.
- During episodes of low liquidity, trades have a higher price impact than under normal conditions.
- The U.S. authorities’ Joint Staff Report (JSR (2015)) finds that High-Frequency Trading (HFT) firms played a dominant role in Treasury markets during the October 15 event window; during the event window the share of HFT activity increased markedly while HFT firms and dealers reduced liquidity provision, resulting in large price moves.
- The paper complements the Joint Staff Report by analyzing market safeguards during the event and proposing measures to mitigate vulnerabilities in the Treasury market.

---

### I. Changes in the structure of the U.S. Treasury market and their impact on liquidity

### A. Electronic trading and the rise of new market makers
- Technological advances and competition from new entrants reshaped the Treasury market; most trading of Treasury securities now occurs electronically through “execution” strategies and increasingly through “automated” strategies.
  - Execution strategies: use technology to optimally execute an order by choosing venue, slicing orders, and/or using algorithms to minimize price impact.
  - Automated decision-making strategies: algorithms make investment decisions that are executed electronically via execution strategies.
- Treasury futures and benchmark Treasury bonds have become almost exclusively traded on electronic platforms:
  - 2001–2002: around 80 percent of benchmark Treasury bonds traded on electronic platforms (Barclay et al. (2006)).
  - End of 2004: more than 95 percent of inter-dealer trading of active issues occurred on electronic platforms (Jiang et al. (2014)).
  - Treasury Borrowing Advisory Committee (2013): benchmark Treasury bonds and futures are entirely traded on electronic platforms.
- Electronic trading platforms introduced new order execution types (limit orders, market orders, iceberg orders, fill-or-kill orders), increasing competition with traditional market makers.
- The inter-dealer trading venues (examples cited) and the CME for Treasury futures support electronic trading; some platforms have opened to non-dealers.
- HFT’s market share (estimates cited in Box 1 and Box 2):
  - HFT is estimated to account for more than half of trading activity on benchmark Treasury bonds and Treasury futures.
  - Cash market HFT activity accounts for 40 to 50 percent of trading volumes (multiple studies cited).
  - Futures market HFT activity estimated at more than 60 percent of volumes (Tabb (2012); CFTC (2013) Concept Release).
  - Reported share on BrokerTec among top 10 trading firms: dealers 14 percent; HFT firms 86 percent (Smith (2015)).

Box 1 (Liquidity Consumption and Provision) — key points preserved
- Market liquidity definition: ability to sell a security quickly, in sufficient quantity, with low cost and low price impact.
- Liquidity provision: traders inserting limit orders (“passive” orders).
- Liquidity consumption: traders inserting market orders (“aggressive” orders); market orders require immediacy and can have large price impact when liquidity provision is low.
- Order book: repository of outstanding limit orders; market depth measures total number of contracts offered on either side.
- Stop orders: hidden orders that appear in the order book when a threshold is breached.
- HFT estimates repeated: HFT accounts for 40–50 percent of cash market volumes; >60 percent in futures; other cited estimates and references noted.

Box 2 (Definition of High Frequency Trading (HFT)) — key points preserved
- HFT is a subset of automated (algorithmic) trading strategies characterized by: i) proprietary trading; ii) very short holding periods; iii) submission of large number of orders quickly modified or cancelled; iv) proximity to the trading platform to minimize latency.
- Reactive Trades: trades occurring just after an order book update — more likely to be HFT.
- Measurement challenges: no consensus on HFT measurement; direct approaches (firm business nature), indirect approaches (trading/quoting patterns, order-to-trade ratio, intraday inventories, lifetime of orders, strategic runs), or identifying HFT strategies.
- Empirical constraints: commercial data often lack participant identity; supervisory data confidentiality limits access.

### B. Changes in traditional market makers’ business models
- Bank-based broker-dealers shifted from risk-warehousing (inventory-based market making) toward risk distribution (broker role matching buyers and sellers without significant inventory).
- Under risk-distribution, banks act as pass-through agents, sending client orders directly to the market and reducing balance sheet risk exposure.
- Consequences of the shift:
  - Reduced shock-absorbing pools of liquidity (fewer bank inventories available to absorb large trades).
  - Tradable risk circulates more quickly, potentially increasing the frequency of liquidity shortages and volatility spikes.
- Drivers of the shift:
  - Technology (electronic trading platforms enabling efficient matching and automated strategies).
  - Post-GFC business-model changes: banks refocusing on more profitable activities and reduced risk appetite.
  - Evidence: Adrian et al. (2013) — during the 2013 “taper tantrum” dealers with more regulatory capital capacity sold more of their inventories than those with tighter capital constraints, suggesting risk appetite affects warehousing decisions.
- Regulatory changes and capital/liquidity requirements:
  - Market participants point to the Leverage Ratio (LR) and Supplementary Leverage Ratio as raising the capital cost of high-volume/low-margin activities such as market making.
  - The Net Stable Funding Ratio may raise the relative cost of short-term funding and affect market making going forward.
  - Changes in accounting (FASB 166 and 167) and Basel III adoption will raise capital standards and alter reported leverage, potentially reducing spare balance sheet capacity.
- Monetary accommodation effects:
  - Prolonged low volatility and low expected risk make broker roles more profitable than risk warehousing.
- Observed changes in primary-dealer behavior:
  - Primary dealers reduced holdings of U.S. Treasuries since mid-2013 (figures referenced).
  - Repo operations declined.
  - Turnover of U.S. Treasuries for primary dealers fell to less than 5 percent, compared to 10 to 12 percent before the GFC.

### C. Implications for market liquidity
- Technology, competition, and banks’ shift to risk distribution have led to a longer-term reduction in liquidity provision by traditional market makers.
- Conventional measures of liquidity may overstate available liquidity because displayed liquidity can disappear quickly (“illusion of liquidity”).
- In low-liquidity episodes, market orders can have outsized price impact; such jumps can trigger further trading and create positive feedback loops amplifying initial shocks (reference to IMF (2015b) and other episodes: Swiss Franc January 2015; Bund volatility April–May 2015; equity flash crash August 24, 2015).

---

*Source: _wp15222 - Introduction (excerpt) from the provided IMF PDF content.*

### 2. Amount of Repo Financing

### 2. Amount of Repo Financing

### C. Implications for market liquidity
- Changes in market structure and the rise of electronic market makers have transformed liquidity provision in the Treasury market; traditional and new market makers operate with leaner inventories, affecting their capacity to make markets during volatile periods (IMF (2015a)).
- Market makers may be unwilling to provide liquidity due to lower risk appetite and lower incentives (no contractual obligation with trading venues to make markets and less client accountability and reputational risk due to the anonymity of trading).
- ECB survey findings:
  - The proportion of respondents indicating a limited ability to make markets in times of stress increased from 11 to 33 percent between December 2013 and 2014 (Figure 2.1).
  - The increase was mostly due to the lack of willingness to take on risk (Figure 2.2).
  - During the “taper tantrum”, 43 percent of respondents reduced their market making activity due to lower willingness to take on risk, compared to 29 percent due to risk management constraint.9
- Electronic market makers:
  - Raman et al. (2012) show electronic market makers reduce liquidity provision significantly when volatility is high; professional traders tend to increase liquidity in the face of stress owing to a longer trading horizon.
  - Electronic trading allows considerably shorter trading horizons and anonymity, reducing incentives to provide liquidity under stress due to lack of reputational effects.
  - Korajczyk and Murphy (2014) show HFT firms employing market making strategies on Canadian equities reduced liquidity provision to large institutional trades by 42 percent during stress episodes while liquidity provision by designated market makers remained unchanged.
- Survey perceptions:
  - ECB respondents see growing use of HFT and automated trading as not a consistent source of liquidity during times of market stress; under unfavorable market conditions HFT and automated trading are seen as likely to amplify volatility and lessen market liquidity (ECB (2015)).
  - Non-bank financial institutions are generally seen as liquidity takers under stressed market conditions because of their lack of market making obligations and low commitment to their client base.

*Italicized source attribution: IMF staff summary based on the supplied content.*

---

### II. A CASE STUDY: THE OCTOBER 15, 2014 “FLASH RALLY”

#### A. The events of October 15, 2014: a historical perspective
- On October 15, 2014, U.S. Treasuries experienced one of the largest intraday changes in yields of the last 25 years.
  - Yields on the 10-year bond fell 37 basis points during the morning session to an intraday low of 1.86 percent before rebounding quickly (Figure 3.1).
- The entire Treasury bond and futures curves were affected, with larger price moves for longer-dated tenors such as the 30-year bond; movements were larger on only three other occasions, all driven by significant policy announcements (Figure 3.2).
- Trading volumes:
  - Traded volumes on BrokerTec surged to US$946 billion on October 15, more than 40 percent higher than the previous record of US$662 billion during the May 2013 “taper tantrum”.
- Market conditions around the retail sales release:
  - Ahead of the retail sales report, there were around 25,000 contracts offered for the 10-year Treasury futures at the top 10 prices, close to the average observed on previous days.
  - Liquidity vanished shortly before the 8:30 data release; the data were modestly worse than consensus.
  - The typical rebound of liquidity after the release did not progress as usual: liquidity recovered for about 20 minutes but, from 8:50, started to decline rapidly through 9:30 when total market depth was less than 20 percent of the pre-announcement level.
  - Similar trends observed in the cash market.
- Impact on related asset classes:
  - Interest rate swaps fell 30 basis points between 8:00 and 9:40; equity markets experienced a significant decline before rebounding quickly.
  - Equities, foreign exchange, and commodities did not exhibit large price moves; between 9:30 and 9:40: gold increased by 0.8 percent, the euro appreciated by 0.9 percent against the dollar, and yields on 10-year German bonds declined by 4 basis points against 15 basis points for 10-Year U.S. Treasuries.
  - The lack of large spillovers is prima facie evidence that the flash rally was not driven by fundamentals.
- Post-event liquidity:
  - Following the event, market depth fell by 40 percent in Treasury bonds and by more than 50 percent in Treasury futures (Figure 5).
  - Since then market depth has partially recovered in futures and cash markets but remains below long term averages.

#### B. The events of October 15, 2014: a tick-by-tick analysis
- Phases of price action after the 8:30 retail sales report:
  - (i) 8:30–9:33: steady rise in Treasury futures prices (falling yields).
  - (ii) 9:33–9:40: sudden sharp increase in prices.
  - (iii) 9:40–10:00: sharp decline in prices until market settled into more normal trading amid low liquidity.
- Liquidity evolution:
  - Market depth declined in consecutive steps from around 20,000 contracts offered at the 10 best prices to about half at 9:30.
  - The erosion of liquidity enabled the flash rally to take off at around 9:33 with a sudden price increase and further declines in market depth.
- Large trades and their impact:
  - A series of large aggressive buy transactions occurred at 9:33:45, including the largest buy trade on that day for 1,000 contracts (Figure 6.3, red circle).
  - These transactions consumed about 10 percent of the market depth available at the 10 best prices (Figure 7.1).
  - Liquidity depleted further to less than 5,000 contracts at the 10 best prices (Figure 6.2).
  - Bid-ask spreads remained tight because offers existed at each price level, but quantities offered were very small.
- Reversal mechanics:
  - At around 9:40, large sell limit orders for 30-year Treasury bonds and futures became visible, providing liquidity to buyers and creating a resistance level.
  - In futures, sell orders for more than 1,000 30-year contracts represented 40 percent of the 2,500 contracts offered to buyers at the 10 best prices (Figure 7.2).
  - In the cash market, sell limit orders for US$8 million Treasury bonds represented more than 70 percent of the bonds offered to buyers.
  - Those sell orders were never executed against, but their visibility prompted prices to decline; by 9:45 prices returned to their 9:33 levels amid very low liquidity and record high trading volumes, with short-lived volatile episodes around 9:46 and 9:56.
- Data and note:
  - Analysis primarily based on futures data; BrokerTec cash data yield similar results but cannot be presented due to copyright issues.11,12,13

#### C. Market safeguards
- CME safeguard mechanisms:
  - (i) post-trade risk control ("velocity logic" functionality) that halts trading if price changes over one second exceed a threshold.
  - (ii) pre-trade risk control ("stop logic" functionality) that halts trading if the execution of a stop order would result in large price moves.
- On October 15, safeguards did not trigger because:
  - Prices moved up rapidly but continuously between 9:33 and 9:40, without discrete gaps (Figures 8 and 9).
  - There was no execution of a single stop order that would have resulted in large price moves.
  - The smoothness of the price increase suggests contracts were offered to buyers at each price level, although overall liquidity was scarce and some large trades were executed at multiple price levels.

*Italicized source attribution: IMF staff summary based on the supplied content.*

---

### III. WHAT FACTORS CONTRIBUTED TO THE FLASH RALLY?

#### A. Repositioning by hedge funds
- Background:
  - The Federal Reserve had signaled earlier in 2014 it would taper purchases of U.S. Treasuries and mortgage-backed securities; expectations of approaching normalization of U.S. policy rates built a consensus.
  - Leveraged funds had been steadily reducing net long positions in interest rate futures and assumed substantial net short positions to profit from rising rates.
  - In the weeks prior to October 15, that consensus was questioned due to discouraging global growth data and reports of Ebola cases in the U.S., supporting the possibility of ‘safe haven’ flows into U.S. Treasuries.
  - U.S. Treasury yields declined 15 to 19 basis points across the 2- to 10- year maturities during the first two weeks of October.14
- Hedge fund actions on October 15:
  - Many hedge funds might have had to reduce short positions to avoid further losses; on the morning of October 15, hedge funds covered their short positions by buying Treasuries.
  - A Federal Reserve (FRB (2015)) dealer survey found around two-thirds of respondents indicated hedge funds were net buyers of Treasuries between 8:30 and 10:00 AM (Figure 10.1).
- Limitations of the hedge-fund-repositioning explanation:
  - Most hedge funds’ repositioning in U.S. Treasury futures occurred in September (Figure 10.2), without significant price impact.
  - The flash dynamics occurred more than one hour after the retail sales report, well after the normal time frame for news-driven market reaction.15
  - Repositioning was directional (buying pressure) and thus cannot explain the sharp reversal observed around 9:40 AM.

*Italicized source attribution: IMF staff summary based on the supplied content.*

### 2.  Leveraged  Funds:  Net  Positions  in  U.S.  Treasury

### 2. Leveraged Funds: Net Positions in U.S. Treasury Futures

### B. Withdrawal of liquidity providers — findings
- Market depth was severely reduced during the event as liquidity providers did not replenish consumed liquidity.
- Liquidity provision proxied by order book depth at different price levels:
  - HFT firms typically provide liquidity close to the best prices; liquidity deeper in the order book (away from the best prices) is more likely from dealers.
  - On October 15, the reduction in market depth was higher for orders deep in the order book (6 to 10 best prices), compared to orders close to best prices (1 to 5 best prices).
  - Around 9:10, depth of deep orders declined more than for orders close to the best prices.
- Indirect evidence suggests dealers reduced liquidity provision more than HFT firms during the event.
- Dealer risk warehousing practices may have amplified price swings:
  - Dealers typically internalize a significant share of client orders (risk warehousing) rather than sending orders to the cash market.
  - When faced with significant buying pressure, dealers might have reduced internalization and instead sent client orders to the market, amplifying buying pressure.
- Evidence from consecutive trades:
  - On control days (pre-October 15), there was an average of 20 consecutive trades on the same side (buy or sell).
  - On October 15, this figure declined to less than 10, consistent with reduced warehousing of client orders and dealers potentially “dumping” client orders directly to the interdealer market.
- Comparable behavior was observed during the May 6, 2010 equity flash crash (SEC-CFTC (2010)).

### C. High frequency trading — findings and metrics
- Several proxies indicate higher HFT activity during the event.
- Order book updates:
  - Order book updates surged from 10,000 updates per minute before the event to a peak of 30,000 updates around 9:45 AM on October 15.
- Trade sizes:
  - Trading during the event occurred in unusually small clips; around half of the trades were for only one contract, more than 20 percentage points higher than on October 2.
  - Note: 10-Year Treasury futures contracts have a notional value of US$100,000.
- Reactive trades (RTs) as a proxy for HFT activity:
  - RTs are defined as market orders executed in reaction to an update of the order book and constitute a subset of HFT strategies; estimates are a lower bound of HFT activity.
  - The estimated share of RTs increased after 9:00 on October 15 concurrently with the gradual depletion of market depth.
  - At the peak, more than 16,000 contracts are estimated to be RTs between 9:39 and 9:40, more than 25 percent of the entire volume traded during that period.
- Interaction effects:
  - High HFT activity combined with low market depth likely amplified price dynamics.
  - High HFT activity alone does not necessarily produce significant price changes (spikes on October 8, 16, and 22 are examples), but the combination of withdrawal of liquidity by dealers and aggressive HFT activity amid low depth produced flash dynamics.
  - Positive feedback loop described: Low Market Depth → High Number of Trades → High Price Impact and High Number of Order Book Updates → High HFT Activity → further depletion of market depth.

### D. Authorities’ analysis of the event — key points
- U.S. authorities’ report (JSR (2015), released on 13 July 2015) identified sharp decline in liquidity and the role of HFT activity as main contributors to the flash rally.
- Granular data in the report shows a significant reduction in market depth during the event attributed to HFT firms and dealers reducing liquidity provision:
  - HFT firms reduced quantities offered; broker-dealers increased spreads and removed offers for very brief periods (fraction of seconds).
  - During the event, the share of trading done by HFT firms increased markedly to about 70 percent of trading activity (against 50 percent on control days).
  - HFT firms were aggressively buying during the price rise and selling during the decline; report confirms positive feedback loop with HFT firms (and algorithms) trading with each other very quickly and on a large scale.
- Operational and conduct concerns noted:
  - Surge in message traffic to CME matching engines resulted in increased latency; exchange IT systems could be exposed to operational risk affecting orderly trading.
  - Large number of self-trades during the event (and to a lesser extent on control days) raises concerns about potential manipulative conduct:
    - During the event there was a surge in the amount of self-trades, reaching up to 33 percent of trading volumes in the 5-year note.
    - The Joint report states “the report is not making any findings on the legality of any self-trading that occurred on the days covered in this analysis” (JSR (2015), footnote 28).
- JSR (2015) recommendations summarized:
  - Regulation, inter-agency oversight and monitoring, and risk management practices of the Treasury market should be addressed.
  - Supports a review of the current regulatory framework for Treasury securities and a study of potential registration requirements for HFT firms active in the Treasury market.
  - Echoes TMPG (2015) initiative on reinforcing best practices in the Treasury market.

### IV. Measures to strengthen the U.S. Treasury market — overview
- Three broad areas for consideration:
  - (i) providing incentives for liquidity provision,
  - (ii) improving market safeguards,
  - (iii) enhancing the regulatory framework of the Treasury market.

### A. First line of defense: provide incentives for liquidity provision — recommendations
- Introduce mechanisms to ensure permanent provision of liquidity to prevent flash events and steady replenishment of liquidity consumed.
- Designated market makers (DMMs):
  - Authorities and operators could consider introducing DMMs on the Treasury cash and futures markets.
  - DMMs have contractual obligations to provide liquidity for a list of securities on a trading venue, in exchange for privileges (such as financial incentives), and can supply liquidity on a continuous basis.
  - Quotation from Bank of England publication: “properly calibrated DMM schemes can help prevent the virtuous circle of efficiency and liquidity from degrading into a vicious circle of mispricing and illiquidity” (Benos and Wetherilt (2012)).
  - Note: Unlike equity markets, cash and futures markets for U.S. Treasuries do not have DMMs; CME uses DMMs only for new products.
- Market-making requirements and incentives:
  - EU MiFID revisions introduce requirements for market making strategies: if a firm posts two-way quotes 50 percent of the daily trading hours, it is required to have a market making agreement with the trading venue; agreements should specify minimum obligations and exceptional circumstances where obligations are waived (ESMA (2015a), ESMA (2015b)).
  - Trading venues could consider maker/taker fee structures: require liquidity consumers to pay a fee and offer liquidity providers a rebate.
    - Empirical studies on maker/taker fees on equity markets show mixed effects on market quality, especially when markets are fragmented; more in-depth analysis warranted.
- Market making requirements for primary dealers:
  - U.S. authorities could introduce market making requirements for primary dealers (PDs), as is common in many countries.
  - The Federal Reserve Bank of New York could introduce requirements that primary dealers maintain a significant continuous presence in secondary markets (quoting obligation and/or active participation).
  - Possible obligations: two-way quotes, minimum limits on bid-ask spreads, minimum amounts of quoted volumes, or minimum presence during usual trading periods.
- Waivers and opt-out clauses:
  - Market making requirements should include waivers to avoid forcing market makers to “catch a falling knife” during periods of high stress.
  - Waivers need careful structuring to avoid all-or-nothing triggers that allow mass opt-outs and sudden market depth decreases.
  - Some countries (e.g., UK) have opt-out clauses in place; in the UK PDs are expected to provide two-way quotes between 8:00 AM and 5:00 PM and obligations are suspended temporarily over economic news or data releases (AFME (2014)), though such opt-outs may undermine effectiveness during stress periods.

*Italicized source: IMF staff chapter excerpt titled "2.  Leveraged  Funds:  Net  Positions  in  U.S.  Treasury Futures" from the provided PDF.*

### Box 3: Market Making Obligations for Primary Dealers

### Box 3: Market Making Obligations for Primary Dealers

### Market making obligations for Primary Dealers (PDs)
- In many countries, the privileges of Primary Dealers are tied to market making obligations on secondary markets.
- In most European countries and emerging economies, Primary Dealers (PD) have detailed obligations to either provide two-way quotes continuously or to participate actively in secondary markets.
- Sweden: obligations are compensated by commissions from the Debt Management Office for participating in primary and secondary markets (CGFS (2014)).
- In all countries with obligations, safeguards allow primary dealers to manage their risk by interrupting their liquidity provision and compliance with their obligations is monitored.
- By contrast, primary dealers in the United States, Japan and Germany have currently no obligations to provide liquidity in secondary markets. These three markets are among the most liquid government bond markets in the world but have experienced episodes of significant intraday volatility recently:
  - Japan: Value at Risk shock in April 2013.
  - United States: Flash Rally on October 15, 2014.
  - Germany: Bund turbulence in April-May 2015.

- Countries with quoting obligations for Primary Dealers:
  - Belgium
  - Denmark
  - France
  - Greece
  - India
  - Ireland
  - Italy
  - Malaysia
  - Mexico
  - Netherlands
  - Portugal
  - Spain
  - Sweden
  - United Kingdom

- Countries with active participation obligations for Primary Dealers:
  - Austria
  - Canada
  - Finland

- Countries without secondary market obligations for Primary Dealers:
  - Germany*
  - Japan
  - United States

- Note: *Germany does not have a Primary Dealership system per se but has rules that apply to the Bund Issues Auction Group.

### Second line of defense: strengthen market safeguards
- A market halt triggered by market safeguards may be beneficial to avoid disorderly trading. The CFTC and the SEC (2010) noted that a 5-second halt trading “can be an effective way of providing time for market participants to reassess their strategies, for algorithms to reset their parameters, and for an orderly market to be re-established” (CFTC-SEC (2010)).
- Existing market safeguards could be recalibrated (such as by using a longer window for the “velocity logic” functionality).
- Changes in market depth could be included as one of the inputs for triggering market safeguards to complement current systems which focus on price variation.
- Market safeguards should be introduced in the Treasury cash market. The two main interdealer electronic trading platforms for U.S. Treasury bonds, BrokerTec and eSpeed, do not have any market safeguard mechanism in place.
- Market safeguards should be harmonized with the Treasury futures market to ensure consistency across asset classes, but common safeguards must be carefully calibrated to prevent trading halts from being too frequent and reducing price discovery.
- Italy: secondary market performance of PDs is assessed through the Quotation Quality Index, an indicator based on high-frequency snapshots of the order book for each bond (AFME (2014)).

### Improve the regulatory framework of the Treasury market
- Quote from SEC Commissioner Aguilar (2015): “The Treasury market, the largest and most liquid government securities market in the world, is now arguably the least transparent fixed income market in the U.S.”
- Structural changes in market structure suggest review of current regulatory requirements is needed, including:
  - (i) changes to the regulation, surveillance and supervision of the Treasury market; and
  - (ii) changes to the supervision of market participants in the Treasury market.

- Current regulatory, surveillance and supervision features:
  - Under the Government Securities Law Act of 1986, the Department of the Treasury can promulgate rules governing transactions in government securities by broker-dealers.
  - The SEC, the Financial Industry Regulatory Authority (FINRA) and federal bank regulators can issue sales practices rules for U.S. Treasuries in secondary markets, in consultation with the Department of the Treasury (Department of the Treasury et al. (1998), JSR (2015)).
  - Surveillance since 1992 has been conducted by an interagency working group: U.S. Treasury, U.S. Federal Reserve Board of Governors, Federal Reserve Bank of New York, SEC and CFTC.
  - In the Treasury futures market, the CFTC and the CME, as a self-regulatory organization, are in charge of regulation and supervision.
  - Strong interagency coordination, eventually under the auspices of the Financial Stability Oversight Council, is warranted.

- Regulatory gaps and implications:
  - In cash markets, trading venues and market participants trading Treasuries are not subject to specific requirements, complicating efforts to analyze and contain systemic risks.
  - The SEC and FINRA regulate and supervise the broker-dealers operating Treasury trading platforms (Alternative Trading Systems (ATSs)). However:
    - ATSs trading exclusively Treasuries are exempted from ATS requirements, as Treasuries are not under the scope of Rule 300(a) of the SEC’s Regulation ATS.
    - Granular FINRA rules such as Rule 2121 Fair Prices and Commission do not apply to Treasury securities.
    - Consequently, ATSs are not subject to any requirement regarding market safeguards or liquidity provision schemes.
  - Market participants trading on ATSs are supervised if they are broker-dealers or banks, but other participants — in particular proprietary trading firms — are not supervised directly, even if their actions have systemic implications (operational risk, market manipulation or inadequate risk management practices).

- Recommendations on regulatory perimeter and data:
  - The regulatory perimeter should be expanded: electronic trading platforms for Treasuries should be subject to specific regulations and participants should be supervised to ensure a level playing field and close regulatory loopholes.
  - A consolidated tape of transactions in the Treasury market would help to enhance market surveillance. While all transactions on equities and corporate bonds are reported to FINRA, there is no reporting requirement for Treasury securities.

### Regulation and supervision of HFT and proprietary trading firms
- Regulation of HFT firms has lagged behind changes in the microstructure of the Treasury market.
- Rise of electronic proprietary trading firms has sometimes been favored by regulatory changes; some of those trading firms are not under the regulatory framework in the United States.
- Concerns include market integrity and the need for adequate risk management systems and controls to avoid positive feedback loops.
- Currently, only a limited number of proprietary trading firms are directly supervised, leaving room for regulatory arbitrage.
  - In the Treasury futures market, trading firms are supervised by the exchange and the CFTC if they are registered as Futures Commission Merchants (FCMs).
  - If not registered as FCMs, it is up to the FCMs they trade with to set up adequate risk controls; weak pre-trade risk control requirements for FCMs can create incentives for customers to choose FCMs with the least risk controls (Clark and Ranjan (2012)).
- SEC proposal (SEC (2015)): extend supervision to proprietary trading firms by requiring proprietary trading firms to become members of FINRA.
- The Joint Staff Report (JSR) suggests studying specific registration requirements for HFT firms trading Treasuries.

### Conclusion: causes of the October 15, 2014 flash rally and policy implications
- The flash rally of October 15, 2014 illustrated the fragility of the U.S. Treasury market.
- Key mechanisms and dynamics identified:
  - Erosion of liquidity following a news release allowed individual trades to have an abnormal market impact, setting conditions for flash dynamics.
  - Retail sales data release was widely interpreted as implying the Federal Reserve might raise policy rates later than expected, leading leveraged funds to cover short Treasury positions.
  - Longer-term trend of traditional market makers retreating from warehousing risk reduced stabilizing capacity; other market participants shied away from executing large orders fearing inability to exit positions.
  - Result: amount available for purchase or sale at or near market price fell significantly below normal levels, allowing relatively small orders to move price quickly through successive price levels.
  - Market participants remained active at high frequencies, making numerous small trades at rapidly changing prices, driving total trading volumes to record levels.
  - Because prices did not ‘gap’ (i.e., moved through consecutive levels without skipping), market safeguards were not triggered and did not mitigate the shock.

- Structural changes:
  - Technological advances enabled automation of trading strategies, giving rise to HFT firms that rely on small inventories and adjust holdings rapidly in times of stress.
  - Traditional market makers curtailed liquidity provision and adopted automated trading strategies as well.
  - Drivers include competition, changes in business models since the onset of the Global Financial Crisis (GFC), and possibly new regulations.

- Policy prescriptions and measures to improve resiliency:
  - Incentivize or mandate provision of liquidity in the case of primary dealers.
  - Introduce Designated Market Makers.
  - Improve market safeguards, including by taking account of market depth directly.
  - Strengthen regulatory and supervisory oversight, close regulatory loopholes, and address significant data gaps to give authorities a complete view of the U.S. Treasury market.
  - Design and implement calibrated market safeguards to prevent recurrence.

### Appendix 1: Market safeguards on the Chicago Mercantile Exchange (CME)
- The CME uses two main market safeguards: the Stop Logic and the Velocity Logic functionalities.
  - Stop Logic: a pre-trade risk control designed to limit the price impact of the triggering of a single stop order. If execution of a stop order leads to a large price change above a specified threshold, trading is halted for a few seconds.
  - Velocity Logic: a post-trade risk control designed to prevent excessive price change within a given period of time. If one-second (absolute) price changes are above a threshold, trading is halted for a few seconds.

- Stop Logic numerical example and conditions:
  - Suppose the threshold for the Stop Logic is 6 points.
  - The security is currently traded at the price above 1070, but a seller places a Stop sell order for 250 contracts at the triggering price of 1070.
  - Trade executed at the price of 1070 triggers the stop order to sell 250 contracts at the best price available.
  - Given the low liquidity in the bid-side order book (quantities offered to sellers shown below), the order would be executed down to 7 price levels of the order book (to fill the order of 250 contracts), bringing prices to 1063, which is more than 6 points below the previous price.
  - Because the new price level would be more than 6 points below the current price, the Stop Logic safeguard would halt trading.
  - After a few seconds (typically 5 seconds), trading resumes.

- Stop Logic conditions and order book state (as presented):
  - Stop Logic Value: 6 points
  - Stop sell order: Size of 250 contracts at the triggering price of 1070
  - State of Order book when the Stop sell order is triggered:
    - Price 1070 — Quantities offered to sellers (Buy limit orders): 10 — Cumulated quantities: 10
    - Price 1069 — Quantities: 50 — Cumulated quantities: 60
    - Price 1068 — Quantities: 40 — Cumulated quantities: 100
    - Price 1067 — Quantities: 25 — Cumulated quantities: 125
    - Price 1066 — Quantities: 75 — Cumulated quantities: 200
    - Price 1065 — Quantities: 20 — Cumulated quantities: 220
    - Price 1064 — Quantities: 10 — Cumulated quantities: 230
    - Price 1063 — Quantities: 40 — Cumulated quantities: 270
    - Price 1062 — Quantities: 20 — Cumulated quantities: 290
  - Expected impact: -7 points — Stop Logic will be triggered

- Velocity Logic numerical example and condition:
  - Velocity Logic is designed to detect market movements in either direction larger than a predefined value within a certain time (typically 1.5 point over a one-second window for U.S. Treasury futures).
  - Numerical example:
    - Suppose the Velocity Logic threshold is set to 6 points within a second.
    - Security traded at 1070; large seller orders move market downwards for 7 points in 830 milliseconds.
    - Velocity Logic is triggered and trading is halted for a few seconds.
    - At the end of the trading halt, trading resumes.
  - Velocity Logic condition: Velocity Logic Value — 6 points within one second.

- Design of market safeguards and the events of October 15, 2014:
  - Market safeguards such as Stop Logic and Velocity Logic were not triggered on October 15 because, despite high volatility, prices moved up in small increments over a relatively long period of time.
  - If the velocity logic safeguard had used a longer time window, it would have been triggered. Example: the price change of 10-year UST futures between 9:32 and 9:40 on October 15 was higher than the threshold value of the velocity logic on U.S. Treasury futures (1.5 point).
  - Including the current level of depth in the order book as a trigger for market safeguards could have mitigated the price action on October 15. Market depth was used only indirectly under current settings.
  - Example: an abrupt change in market depth — the 75 percent decline which occurred between 8:50 and 9:30 on October 15 — if included as a trigger for market safeguards, could mitigate ex ante the risk of flash dynamics.
  - Caveat: calibration needs assessment because ahead of news releases market depth typically vanishes and safeguards should not be triggered in those instances.

- Historical triggering of circuit breakers on U.S. Treasury futures (Box 1):
  - Recent events in which the circuit breakers were triggered:
    - 7 June 2013 — 5Y, 30Y, eMini — Non farm payroll
    - 2 August 2013 — 5Y — Non farm payroll
    - 8 November 2013 — 5Y, 30Y — Non farm payroll
    - 10 January 2014 — 5Y — Non farm payroll
    - 4 April 2014 — 5Y — Non farm payroll
    - 5 December 2014 — 30Y — Non farm payroll
  - There were also events where large price movements occurred in hindsight but safeguards were not triggered:
    - 29 February 2012 — 10-year Treasury futures price declined about 0.4 percent within one minute interval around 10:05 am.
    - 23 December 2013 — 30-year Treasury futures price ascended about 4.2 per cent over one minute interval around 2:38 am.
  - In these instances, price movement was smooth at very small intervals (25-millisecond) but very large cumulatively, so safeguards did not trigger.

*Source: IMF Working Paper (content unit: Box 3: Market Making Obligations for Primary Dealers).*

### APPENDIX 2: ESTIMATION OF REACTIVE TRADES

### APPENDIX 2: ESTIMATION OF REACTIVE TRADES

### I. Reactive Trades — definition and motivation
- Reactive Trade (RT) is defined as a trade (market order) placed in reaction to an update of the best quote in the limit order book (Best Quote Update or BQUD).
- RTs are used as a proxy for algorithmic and high frequency trading (HFT) activity.
- Rationale: when a trade occurs very quickly after a BQUD, there is a high probability it comes from an HFT firm (manual traders cannot react as fast).
- Shortcomings of a naive threshold-based approach (e.g., JLV classification using 1 second):
  - During volatile periods, an increased number of BQUDs raises the probability that non-RT trades occur within the threshold by chance, producing upward bias.
  - A 1 second threshold may be too slow for HFT activity, which typically occurs over milliseconds.
- Goal: improve on JLV by modeling occurrence of RTs and non-RTs probabilistically to estimate RT volume more accurately.

### II. Model — events, probabilities, intensities, and trade volumes

- Four mutually exclusive event types during a trading session [time t in 0..T]:
  - Reactive trade (RT)
  - Non-reactive trade (non-RT)
  - Non-simultaneous best quote update (non-simultaneous BQUD)
  - Simultaneous BQUD (BQUD occurring with a trade)

- Observable and unobservable processes:
  - Observed: N_Trade(t) := N_RT(t) + N_non-RT(t)
  - Observed: N_UD(t) = number of BQUDs by time t
  - Unobserved: N_RT(t), N_non-RT(t)
  - N_UD̅(t) := number of non-simultaneous BQUDs by time t
  - Difference N_UD(t) − N_UD̅(t) = number of simultaneous BQUDs up to t

- Probability assumptions for very small Δ>0:
  - Non-RT occurrence:
    - P(N_non-RT(t+Δ) > N_non-RT(t)) = λ_non-RT(t) Δ  (equation (1))
  - Non-simultaneous BQUD occurrence:
    - P(N_UD̅(t+Δ) > N_UD̅(t)) = λ_UD̅(t) Δ  (equation (2))
  - RT occurrence conditional on latest BQUD time τ_UD(t):
    - P(N_RT(t+Δ) > N_RT(t) | τ_UD(t)) = λ_RT(t, τ_UD(t)) Δ  (equation (3))
    - λ_RT is a function of current time and time of latest BQUD: RTs react to past BQUDs through τ_UD(t).
  - Probability of a BQUD conditional on a trade at time t:
    - P(N_UD(t) > N_UD(t−) | N_Trade(t) > N_Trade(t−)) = q(t)  (equation (4))
    - q(t) assumed independent of other information up to t.

- Additivity and time-to-next-event distribution:
  - Define λ(t, τ) := λ_RT(t, τ) + λ_non-RT(t) + λ_UD̅(t).
  - Probability of no event between t and s conditional on τ_UD(t):
    - P(N_Trade(s) = N_Trade(t), N_UD(s) = N_UD(t) | τ_UD(t)) = exp(−∫_t^s λ_RT(u, τ_UD(t)) du − ∫_t^s λ_non-RT(u) du − ∫_t^s λ_UD̅(u) du)  (equation (5))
  - Define δ(t) := inf(h≥0; N_Trade(t+h)+N_UD(t+h) > N_Trade(t)+N_UD(t))
  - Equivalent form:
    - P(t + δ(t) > s | τ_UD(t)) = exp(−∫_t^s λ(u, τ_UD(t)) du)  (equation (6))
  - Density of δ(t):
    - f_δ,t(h | τ_UD(t)) = λ(t+h, τ_UD(t)) exp(−∫_t^{t+h} λ(u, τ_UD(t)) du)  (equation (7))

- Specification of intensities (piecewise and parametric forms):
  - Divide observation period 0 = t_0 < t_1 < ... < t_n = T.
  - Assume λ_non-RT(t) and λ_UD̅(t) are constant within each interval (t_{i−1}, t_i]:
    - λ_non-RT(t) = λ_non-RT_i for t∈(t_{i−1}, t_i]  (equation (8))
    - λ_UD̅(t) = λ_UD̅_i for t∈(t_{i−1}, t_i]  (equation (9))
  - Assume separable form for λ_RT:
    - λ_RT(t, τ) = λ_RT_i g(t − τ) for t∈(t_{i−1}, t_i]  (equation (10))
    - g: continuous on [0,+∞), g(0)=1, g(h)>0 for h>0, g(h)→0 as h→+∞.
  - Specific functional form chosen for g for tractability:
    - g(h) = exp(−h^2 / 2σ^2)  (equation (11))
  - Probability q(t) assumed piecewise-constant:
    - q(t) = q_i for t∈(t_{i−1}, t_i]  (equation (13))

- Volume per trade specifications:
  - Number of contracts per RT, V_RT, and per non-RT, V_non-RT, assumed exponentially distributed:
    - V_RT(t) | event is RT ∼ Exponential(η_RT)  (equation (14))
    - V_non-RT(t) | event is non-RT ∼ Exponential(η_non-RT)  (equation (14))
  - Justification: Q-Q plots of October 15 volumes vs exponential show fit for trades below 20 contracts (95 percent of all trades).

### III. Estimation of parameters and RT volumes

- Parameters to estimate:
  - Θ = { σ, (λ_RT_i, λ_non-RT_i, λ_UD̅_i, q_i)_{i=1,...,n} } (frequency parameters)
  - η_RT and η_non-RT (volume/exponential parameters)

- Likelihood construction using observed events:
  - Observe events at times 0 ≤ t_1 < t_2 < ... < t_J ≤ T.
  - Define inter-event times h_1 = t_1, h_j = t_j − t_{j−1} for j=2,...,J.
  - Define indicator X_j = I{N_Trade(t_j) > N_Trade(t_j−)} = 1 − I{N_UD̅(t_j) > N_UD̅(t_j−)} (trade at t_j).
  - Define indicator Y_j = I{N_UD(t_j) > N_UD(t_j−)} (BQUD at t_j).
  - Conditional probabilities:
    - P_tj(X_j = 1) = {λ_RT(t_j, τ_UD(t_j)) + λ_non-RT(t_j)} / λ(t_j, τ_UD(t_j))
    - P(Y_j = 1 | X_j = 1) = q(t_j)
    - P(Y_j = 1 | X_j = 0) = 1
  - Density of inter-event time conditional on past: f_δ,t_{j−1}(h_j | τ_UD(t_{j−1})) from equation (7).
  - Log-likelihood l({h_j, X_j, Y_j, τ_UD(t_j)}_{j=1,...,J} | Θ):
    - l = log f_δ,0(h_1 | τ_UD(0)) + Σ_{j=2}^J log f_δ,t_{j−1}(h_j | τ_UD(t_{j−1}))  (time observed between events)
      + Σ_{j=1}^J (log P_tj(X_j) + log P_tj(Y_j | X_j))  (types of events observed)  (equation (15))
  - Estimate Θ by maximum likelihood.

- Specification of time intervals and smoothing:
  - Trade-off: too small n → cannot capture time variation; too large n → overfitting.
  - Practically estimate parameters over one-minute intervals and smooth estimates using a Kalman filter.
  - Observation error matrices at each i = 1,...,n set at inverse of estimated Fisher information matrix.
  - Robustness: checks with 5-minutes intervals yielded similar results.

- Validation via simulation (one-minute period reflecting October 15 estimates):
  - Parameters set in simulation: σ = 0.02, λ_RT = 300, λ_non-RT = 60, λ_UD̅ = 30, q = 0.3.
  - Generated 100 simulated datasets; estimated parameters for each.
  - Table 1 summary of simulation results (means and standard deviations of estimated parameters, simulation # = 100):
    - σ: True Value 0.02; Mean of Estimated Parameters 0.02033; Standard Deviation 0.00091
    - λ_RT: True Value 300; Mean of Estimated Parameters 296.0; Standard Deviation 8.8
    - λ_non-RT: True Value 60; Mean of Estimated Parameters 59.29; Standard Deviation 8.78
    - λ_UD̅: True Value 30; Mean of Estimated Parameters 29.91; Standard Deviation 0.70
  - Conclusion: estimation method performs well when data are generated from the model.

- Estimation of RT volume and posterior classification
  - Let π(t) = λ_RT(t, τ_UD(t)) / (λ_RT(t, τ_UD(t)) + λ_non-RT(t)) be prior probability that an observed trade at time t is an RT.
  - Density of trade volume V(t):
    - f_V(t)(v) = π(t) η_RT exp(−η_RT v) + (1 − π(t)) η_non-RT exp(−η_non-RT v)  (equation (16))
  - Posterior probability that a trade at t is RT given observed volume V(t):
    - π_post(t) = [π(t) η_RT exp(−η_RT V(t))] / [π(t) η_RT exp(−η_RT V(t)) + (1 − π(t)) η_non-RT exp(−η_non-RT V(t))]  (equation (17))
  - Aggregate volumes between t and s:
    - A V_RT(t, s) = Σ_{h ∈ (t, s], N_Trade(h) > N_Trade(h−)} π_post(h) V(h)  (equation (18))
    - A V_non-RT(t, s) = Σ_{h ∈ (t, s], N_Trade(h) > N_Trade(h−)} (1 − π_post(h)) V(h)  (equation (19))
  - Use AV_RT and AV_non-RT to compute share of RT activity as a proxy for some HFT strategies.

*Source: APPENDIX 2: ESTIMATION OF REACTIVE TRADES*

### REFERENCES

### REFERENCES

### Research on high-frequency trading, electronic trading, and market making
- Adrian, T., M. Fleming, J. Goldberg, M. Lewis, F. Natalucci and J. Wu (2013), “Dealer balance sheet capacity and market liquidity during the 2013 selloff in fixed-income markets”, Liberty Street Economics, 16 October. 
- Anand, A. and K. Venkataraman (2012), “Should exchanges impose market maker obligations?”, available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2179259 .
- Barclay, M., T. Hendershott, and K. Kotz (2006), “Automation versus Intermediation: Evidence from Treasuries Going off the Run.” Journal of Finance. Vol. 61, 2395–2414.
- Benos, E. and A. Wetherilt (2012), “The role of designated market makers in the new trading landscape”, Bank of England Quarterly Bulletin 2012Q4.
- Bouveret, A., C. Guillaumie, C. Aparicio Roqueiro, C. Winkler and S. Nauhaus (2014). “High-frequency trading activity in EU equity markets”, ESMA Economic Report No. 1.
- Brogaard, J. (2010), “High Frequency Trading and its impact on market quality”, mimeo Kellogg School of Management Northwestern University School of Law.
- Brogaard, J., T. Hendershott, and R. Riordan, (2014). “High Frequency Trading and Price Discovery.” Review of Financial Studies, Vol. 27 (8), 2267-2306.
- Hagströmer, B., L. Nordén, and D. Zhang, (2014), “How Aggressive Are High-Frequency Traders?”, Financial Review, Vol. 49 (2), 395-419.
- Hasbrouck, J. and G. Saar, (2013), “Low-latency trading”, Journal of Financial Markets 16, 646-679.
- Jiang, George J., Ingrid Lo and Giorgio Valente (2014), “High-Frequency Trading around Macroeconomic News Announcements: Evidence from the U.S. Treasury Market”, Bank of Canada Working Paper/Document de travail 2014-56.
- Kirilenko, A., A. S. Kyle, M. Samadi and T. Tuzun (2014), “The Flash Crash: The Impact of High Frequency Trading on an Electronic Market”, CFTC Research Paper, Washington D.C.
- Korajczyk, R. and D. Murphy (2014), “High Frequency Market Making to Large Institutional Trades”, available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2567016 .
- Malinova, K. and A. Park (2015), “Subsidizing Liquidity: The Impact of Make/Take Fees on Market Quality”, Journal of Finance, Vol. 70 (2), 509-536.
- Raman, V., M. Robe and P. K. Yadav (2012), “Electronic Market Makers, Trader Anonymity and Market Fragility”, CFTC Research Paper. Washington D.C.
- Saar, A. and T. Lybek (2002), “Measuring Liquidity in Financial Markets”, IMF Working Paper No. 02/232. (Washington: International Monetary Fund).
- Venkataraman, K. and A. Waisburd (2007), “The value of the designated market maker”, Journal of Financial and Quantitative Analysis, Vol. 42, 735-758.

### Market structure, liquidity, and fixed-income market studies
- Cardella, L, J. Hao and I. Kalcheva (2015), “Make and Take Fees in the U.S. Equity market”, available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2149302 .
- Fleming, M. (2009), “The Microstructure of a U.S. Treasury ECN: The BrokerTec Platform”. Federal Reserve Bank of New York Staff Report No. 381.
- Kite, S. (2010). “Algos Take Hold in Fixed-Income Markets.” Securities Technology Monitor, February, available at http://www.highbeam.com/doc/1G1-218450998.html .
- Light, D. (2014a), “The Costs of Transacting in Public Markets (Part 2)”, CrossRate Technologies LLC blog, 25 September 2014, available at http://www.crossratetechnologies.com/uncategorized/the-cost-of-transacting-in-the-public-markets-part-2/ .
- Light, D. (2014b), “Liquidity is not what it used to be”, CrossRate Technologies LLC blog, 20 October 2014, available at http://www.crossratetechnologies.com/high-frequency-trading-in-u-s-treasuries/liquidity-is-not-what-it-used-to-be/.
- Nanex, 2014, “15-Oct-2014: Treasury Flash Crash”, available at http://www.nanex.net/aqck2/4681.html .
- Smith, Robert Mackenzie (2015), “Client list reveals HFT dominance on Brokertec”, Risk Magazine, 23 September 2015.

### Regulatory, supervisory, policy reports, and official studies
- Aguilar, L. (2015), “Ere Misery Made Me Wise-The Need to Revisit the Regulatory Framework of the U.S. Treasury Market”, Public Statement, July 14, 2015.
- Association for Financial Markets in Europe (2014), “European Primary Dealers Handbook”, available at http://www.afme.eu/dynamic.aspx?id=1992 .
- Chicago Mercantile Exchange (2014), “Adoption of Rule 575 (“Disruptive Practices Prohibited”) and Issuance of CME Group Market Regulation Advisory Notice RA1405-5”. August 28, 2014.
- Clark, C. and R. Ranjan (2012) “How do broker-dealers/Futures commission merchants control the risks of high-speed trading?”, Federal Reserve Bank of Chicago Policy Discussion Paper No 2012-2.
- Committee on the Global Financial System (2014), “Market-making and proprietary trading: Industry trends, drivers and policy implications”, CGFS Papers No 52. Basel, Switzerland.
- Commodities Futures Trading Commission (2013), “Concept Release on Risk Controls and System Safeguards for Automated Trading Environments”, 9 September, Washington D.C.
- Department of the Treasury, Securities and Exchange Commission and the Board of Governors of the Federal Reserve System (1998), “Joint study of the regulatory system for government securities”. Washington D.C.
- European Central Bank (2014), “ Survey on credit terms and conditions in euro-denominated securities financing and OTC derivatives markets (SESFOD)”, January.
- European Central Bank (2015), “ Survey on credit terms and conditions in euro-denominated securities financing and OTC derivatives markets (SESFOD)”, January.
- European Securities and Markets Authority (2015a), “Final Report- Draft Regulatory and Implementing Technical Standards MiFID II/MiFIR”, ESMA/2015/1464, Paris.
- European Securities and Markets Authority (2015b), “Regulatory technical and implementing standards – Annex I  MiFID II/MiFIR”, ESMA/2015/1464 Annex I, Paris.
- Federal Reserve (2015), “Senior Credit Officer Opinion Survey on Dealer Financing Term (SCOOS)”, January.
- International Monetary Fund, (2014), “Improving the balance between financial and economic risk taking”, Global Financial Stability Report Chapter one, October. Washington D.C.
- International Monetary Fund, (2015a), “Enhancing policy traction and reducing risks”, Global Financial Stability Report Chapter one, April. Washington D.C.
- International Monetary Fund, (2015b), “Market liquidity-Resilient or fleeting?”, Global Financial Stability Report Chapter two, October. Washington D.C.
- Joint Staff Report (2015), “The U.S. Treasury Market on October 15, 2014”, Joint Staff Report,  U.S. Department of Treasury, Board of Governors of the Federal Reserve System, Federal Reserve Bank of New York, U.S. Securities and Exchange Commission, U.S. Commodity Futures Trading Commission, July 13, 2015. Washington D.C.
- Potter, S. (2015), “Challenges posed by the evolution of U.S. treasury markets”, remarks at the 2015 Primary Dealer Meeting, New York City, April 13.
- Securities and Exchange Commission (2015) “Exemption for Certain Exchange Members”, Proposed rule, 25 March 2015. Washington D.C.
- Securities and Exchange Commission and Commodities Futures Trading Commission (2010), “Findings regarding the market events of May 6, 2010”, Washington D.C.
- Treasury Borrowing Advisory Committee (2013), “Electronic Trading in the Secondary Fixed Income Markets”. Presentation. Washington D.C.
- Treasury Market Practices Group (2015), “Automated Trading in Treasury Markets-White Paper”, 8 April 2015. New York City, New York.

### Industry, advisory, and practitioner analyses
- Devasabai, K. (2014), “The makings of a Treasury meltdown”, Risk magazine, December 2014.
- Gemloc advisory services (2010), “Primary Dealers Systems: Draft Background Note”.
- Clark, C. and R. Ranjan (2012) “How do broker-dealers/Futures commission merchants control the risks of high-speed trading?”, Federal Reserve Bank of Chicago Policy Discussion Paper No 2012-2.
- Brogaard, J. (2010), “High Frequency Trading and its impact on market quality”, mimeo Kellogg School of Management Northwestern University School of Law.
- Tabb, L. (2012), “Written Testimony to the United States Senate Committee on Banking, Housing, and Urban Affairs”, CEO, TABB Group.” September 20.

*Source: _wp15222 - REFERENCES*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp15222.pdf_