## Money Market Fund Growth During Hiking Cycles

## Source details

**Canonical URL:** [Money Market Fund Growth During Hiking Cycles](https://www.imf.org/-/media/files/publications/wp/2025/english/wpiea2025150-source-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2025/english/wpiea2025150-source-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2025/english/wpiea2025150-source-pdf.pdf.json)

---

### Executive Summary — Overview and Key findings
- MMFs are characterized as safe and liquid short-term investment vehicles.
- Global MMF assets rose from around €7 trillion in 2020 to about €11 trillion at the start of 2025.
- MMFs now account for roughly 15% of the world’s open-ended mutual funds.
- Analysis covers MMF growth across nine countries spanning four global regions.
- Central proxy: the MMF yield advantage (MMF spread) — the gap between MMF yields and alternative cash investments such as bank deposits — used to capture investors’ yield-seeking behavior.
- Empirical approach: examine sensitivity of MMF growth to the MMF spread across the policy cycle, controlling for the macro-financial environment and fund characteristics.
- Specific event analyzed: the March 2023 US regional banking crisis and broader 2023 banking turmoil; incremental change in MMF spread sensitivity examined for the three-month period after the turmoil.
- Key findings:
  - MMF yield advantage is a primary determinant of MMF growth; MMFs grow significantly faster when they offer better returns relative to bank deposits.
  - Relationship is especially pronounced during periods of monetary policy tightening.
  - Little evidence for a panic-induced inflow to MMFs in the wake of the 2023 banking turmoil; sensitivity to MMF yields increased in several regions but overall MMF growth was lower or similar to previous hiking periods once controls are included.
  - Fund structure and currency denomination shape investor preferences: public vs private debt composition conditions investor sensitivity to risk and return; offshore USD-denominated MMFs tend to grow faster compared to local currency funds.

### The size of the global MMF sector
- MMFs are open-ended funds investing in diversified portfolios of short-term, safe and liquid assets; typically offer daily redemptions.
- The global MMF sector is about 10.5 trillion euros.
- The global mutual fund sector is about 74 trillion euros.
- The US holds the largest global MMF share, just shy of 60%.
- Europe and China account for approximately one third of the US size.
- Jurisdictions beyond the US, Europe, and China account for less than 2% of the global MMF sector size.
- In jurisdictions with at least 1% of the global MMF share, MMFs make up about 10 percent to 20 percent of the mutual fund sector; exceptions with much larger shares include China and Mexico.
- European MMFs are mainly concentrated in Ireland, France and Luxembourg; Ireland and Luxembourg cater primarily to a mostly non-domestic investor clientele and host many off-shore dollar funds.
- Data correspond to 2025Q2 values.

### Motivation — Yield advantage, 2023 turmoil, and fund structure
- MMF yield advantage as a determinant of MMF flows:
  - MMF yields adjust more swiftly during rate increase cycles; bank deposit rates exhibit stickiness (deposit betas).
  - Investors pivot toward MMFs when monetary policy tightening widens the spread between MMF and bank rates.
  - The MMF spread typically increases during hiking cycles and declines outside these periods; the 2022 hiking cycle drove a strong net increase in the spread.
- Impact of the 2023 banking turmoil:
  - The 2023 turmoil (SVB and Credit Suisse episodes) did not constitute a broad systemic crisis, enabling clearer observation of substitution dynamics between banks and MMFs during a global hiking cycle.
  - Financial stress receded quickly due to swift policy responses.
  - Empirical focus: whether MMF growth during March–June 2023 was significantly higher than in previous hiking cycles, after controlling for macro-financial environment and fund characteristics.
- Structure and types of MMFs as a driver of flows:
  - US: Government MMFs account for more than 80 percent of assets held by US taxable MMFs after 2015 reform; government MMFs stabilized the US sector and saw substantial inflows during Covid, while prime MMFs saw outflows.
  - Outside US: public-debt MMFs are less prominent; public debt funds have about 40 percent share in Korea, less than 20 percent in remaining countries; EU-equivalent government MMFs (CNAVs) account for only 10 percent of total EU MMFs.
  - Private-debt MMFs outside the US tend to offer higher returns, may restrict on-demand access, and often include bank deposits.
  - Currency denomination: U.S. dollar is dominant; Ireland and Luxembourg have around half of AUMs denominated in U.S. dollars; euro-denominated funds concentrated in Europe (France largest).

### Data, coverage, and construction
- Fund-level NAV data: monthly observations from Lipper, January 2004 to September 2023.
- Key dependent variable: year-on-year growth rate of NAVs for individual MMFs.
- Growth rates above 100 percent in absolute value excluded; robustness checks exclude >75 percent and >50 percent.
- MMF spread constructed as cumulative change in (overnight money market rate − national average bank deposit rate) over the last three years; proxy takes both negative and positive values.
- USD (EUR) denominated MMFs outside the USA (euro area) are assigned the US (euro area) overnight rate.
- Monetary policy hiking periods: defined from first hike until month prior to first cut; monthly policy rates from Bloomberg used.
- Control variables: inflation and GDP year-on-year growth rates, VIX, yield curve spread (10y−2y); for jurisdictions without their own yield curves, US spread assigned to EMs and dollar funds, German spread assigned to remaining euro denominated funds.

### Methodology and empirical specifications
- Baseline panel regression (Eq(1a)): Δy_{t,i} = β Δpr_t + controls_{t,i} + FE_{t,c,i} + ε_{t,i}, where Δy_{t,i} is year-on-year NAV growth and Δpr_t is change in policy rate.
- MMF yield advantage specification (Eq(1b)): Δy_{t,i} = β ΔSpread_{t,c} + γ Dummy_{HC_{t,c}} + controls_{t,i} + ε_{t,i}.
- Difference-in-difference during hikes (Eq(1c)): includes Spread * Dummy_{HC}.
- Post-turmoil specification (Eq(2b)): Δy_{t,i} = β (Period_{t,c} * ΔSpread_{t,c}) + γ Period_{t,c} + δ ΔSpread_{t,c} + ζ controls_{t,i,c} + ε_{t,i}, with Period_{t,c} equal to 1 during March–June 2023.
- Fund-type specifications (Eq(3a)–(3c)) test interactions with Public_{t,i} and DollarFund_{t,i} dummies.
- Estimation: fixed effects, robust standard errors clustered at fund level.

### Empirical results — Monetary policy, MMF spread, and fund characteristics
- Impact of monetary policy (Table 1):
  - Policy rate coefficient: 4.133*** (0.125) reported in column (1).
  - Text interpretation: MMFs in sample grow on average by 2.14 percentage points for 1 percentage point increase in the policy rate (table coefficient reported as 4.133***).
  - Spread coefficient (column (2)): 3.418*** (0.128).
  - Column (3): Spread = 1.077*** (0.126); Dummy (hiking cycle) = 1.028*** (0.214); Spread * Dummy = 0.903*** (0.155).
- Controls and size effects (selected coefficients from Table 1):
  - Dummy (size medium): 16.708*** (0.227) in (1).
  - Dummy (size large): 32.058*** (0.279) in (1).
  - Inflation: -2.433*** (0.072) in (1).
  - GDP: 1.106*** (0.060) in (1).
  - VIX: -0.826*** (0.059) in (1).
  - Yield curve (YC): -2.329*** (0.071) in (1).
- Interpretation:
  - Monetary policy shocks positively associated with MMF growth; larger funds grow significantly faster; MMF assets rise with higher GDP and decline with higher inflation and higher VIX; steeper yield curve associated with modest outflows.
- Impact of the 2023 banking turmoil (Table 2):
  - Full sample (column (1)): Dummy post turmoil = -8.873*** (1.051); Spread = 3.381*** (0.172); Dummy post turmoil * Spread = 2.927*** (0.561).
  - Regional results:
    - USA (col (2)): Dummy post turmoil = -24.984** (17.085); Spread = 5.692*** (0.621); Dummy post turmoil * Spread = 11.632** (5.229).
    - Euro area (col (3)): Dummy post turmoil = -1.428 (2.009); Spread = 10.045*** (0.609); Dummy post turmoil * Spread = -2.272* (1.335).
    - Americas (col (4)): Dummy post turmoil = 0.044 (1.041); Spread = 1.193*** (0.178); Dummy post turmoil * Spread = 2.862*** (0.798).
    - Asia-Pacific (col (5)): Dummy post turmoil = -18.709*** (5.615); Spread = 2.506*** (0.502); Dummy post turmoil * Spread = 5.359* (3.017).
  - Interpretation:
    - Sensitivity to MMF spreads increased in most regions during March–June 2023, indicating intensified yield-seeking.
    - In the U.S., MMF growth was lower in the post-turmoil period once controls are included, but responsiveness to spread rose.
    - Euro area shows an insignificant post-turmoil growth dummy and a negative Spread interaction at the 10% level.
    - No uniform “flight to safety” into MMFs following SVB and Credit Suisse; higher MMF yields helped sustain flows post-turmoil, but headline inflows reflect macro-financial conditions, fund characteristics, and yield differentials.
- Impact of MMF characteristics (Table 3 and Table 4):
  - Full sample (Table 3):
    - Spread: 1.771*** (0.096) in (1); 2.051*** (0.102) in (2); 1.852*** (0.097) in (3).
    - Public dummy: insignificant across specifications (e.g., 0.213 (1.005) in (1)).
    - Spread * Public (column (2)): -1.579*** (0.185) — a 1 percentage point increase in the spread associated with a 1.58 percentage point smaller inflow to public MMFs relative to non-government MMFs.
    - Dummy post turmoil (column (3)): -3.721*** (0.603).
    - Dummy post turmoil * Public (column (3)): -3.922** (1.195).
    - Dollar Fund dummy (offshore USD-denominated funds outside the US): 11.254*** (1.675) in (1) and 11.408*** (1.675) in (2) — offshore USD funds grew about 11 percent faster than other currencies.
  - Regional heterogeneity (Table 4):
    - Spread by region: USA 6.079*** (0.230); Europe 4.913*** (0.220); Europe (alt) 10.680*** (0.301); Americas 2.233*** (0.175); Asia/Pacific 1.750** (0.551).
    - Public dummy by region: USA 5.081*** (1.097); Europe -3.846* (1.786); Asia/Pacific 10.578*** (2.832).
    - Spread * Public interaction: USA -4.136*** (0.281); Europe -2.702*** (0.367); Americas -3.322* (1.438).
    - LVNAVs in Europe: Dummy (LVNAV) = 5.248** (1.744) and Spread * Dummy (LVNAV) = 4.033*** (0.508) in one specification.
    - Dollar Fund effects by region: USA 11.707*** (1.829); Europe 11.333*** (1.792); Americas 13.291*** (0.980).
  - Interpretation:
    - Public debt MMFs do not universally attract higher inflows; when spreads rise investors shift toward non-public MMFs that offer higher yields.
    - Offshore USD-denominated MMFs exhibit a strong growth advantage (~11 percent faster on average).
    - Regional patterns reflect market structure: U.S. public MMFs grew faster on average post-2015 but attract relatively fewer inflows when spreads widen; Europe shows LVNAV strength.

### Synthesis of empirical findings
- Monetary policy tightening:
  - Higher policy rates associated with increased MMF growth.
  - MMF yield advantage (Spread) is a significant driver of flows and its effect is amplified during hiking cycles.
- 2023 banking turmoil:
  - MMF growth overall was lower or similar to prior hiking episodes once controls are included.
  - Sensitivity to MMF spreads increased in most regions during March–June 2023, indicating stronger yield-seeking.
  - No uniform flight to safety into MMFs following SVB and Credit Suisse; headline inflows largely reflect multiple factors.
- Fund characteristics and currency:
  - Larger funds grew significantly faster (concentration effect).
  - Public debt MMFs generally experienced relatively smaller inflows when spreads rose.
  - Offshore USD-denominated MMFs grew about 11 percent faster on average.
- Regional heterogeneity:
  - U.S.: government/Treasury MMFs expanded markedly since 2015 but become relatively less preferred when spreads widen.
  - Europe: LVNAVs grew faster and attracted flows when spreads rose; CNAVs less attractive.
  - Asia-Pacific and Americas show differing preferences driven by local market structure and yield opportunities.

### Conclusion and policy implications
- Key conclusions:
  - The MMF yield advantage is a primary driver of MMF growth, typically emerging during monetary policy tightening cycles.
  - The yield advantage supported MMF inflows following the 2023 banking turmoil, although the turmoil itself did not catalyze MMF flows in a uniform or panic-driven manner.
  - USD-denominated MMFs are preferred in several jurisdictions, underscoring the USD’s dominant role in short-term funding markets.
  - Cyclical growth behavior in MMFs can make rapid reversals of flows more likely in times of stress.
- Policy implications and recommendations:
  - Rapid MMF growth during tightening cycles can amplify systemic risks when flows reverse sharply.
  - Cyclical attractiveness of MMFs relative to bank deposits can accelerate deposit outflows, complicating monetary policy transmission and straining bank funding.
  - Policymakers and supervisors should ensure robust MMF regulation and oversight, including:
    - Maintaining adequate liquidity buffers.
    - Ensuring sound risk management practices.
    - Promoting transparency to help MMFs accommodate rise-and-bust cycles and mitigate systemic vulnerabilities.
  - Ongoing monitoring and adaptive regulation are essential as the landscape of cash alternatives evolves.

*IMF WORKING PAPERS Money Market Fund Growth During Hiking Cycles — Working Paper No: WP/25/150*

### Executive Summary ......................................................................................................

### Executive Summary

### Overview
- Money Market Funds (MMFs) are characterized as safe and liquid short-term investment vehicles.
- Global MMF assets rose from around €7 trillion in 2020 to about €11 trillion at the start of 2025.
- MMFs now account for roughly 15% of the world’s open-ended mutual funds.
- The analysis covers MMF growth across nine countries spanning four global regions.

### Research focus and methodology
- Central proxy: the MMF yield advantage (MMF spread) — the gap between MMF yields and alternative cash investments such as bank deposits — used to capture investors’ yield-seeking behavior.
- Empirical approach: examine sensitivity of MMF growth to the MMF spread across the policy cycle, controlling for the macro-financial environment and fund characteristics.
- Specific event analyzed: the March 2023 US regional banking crisis and broader 2023 banking turmoil; incremental change in MMF spread sensitivity examined for the three-month period after the turmoil.

### Key findings
- MMF yield advantage is a primary determinant of MMF growth:
  - MMFs grow significantly faster when they offer better returns relative to bank deposits.
  - This relationship is especially pronounced during periods of monetary policy tightening.
- The 2023 banking turmoil did not produce broad panic-driven inflows into MMFs:
  - Little evidence for a panic-induced inflow to MMFs in the wake of the 2023 banking turmoil.
  - In the US, MMF growth became more sensitive to superior MMF yields after the turmoil, but MMFs did not grow as much as in previous hiking periods once macro-financial and fund characteristics are controlled for.
  - In most global regions, yield-seeking behavior continued in the three-month period after the 2023 turmoil.
  - Asia-Pacific and Americas regions show similar post-turmoil sensitivity patterns; euro area showed a moderating of previously elevated MMF spread sensitivity.
- Fund structure and currency denomination shape investor preferences:
  - Public vs private (non-public) debt composition conditions investor sensitivity to risk and return.
  - Public MMFs dominate the U.S. market, supported by the depth and liquidity of the U.S. Treasury market; outside the U.S., private debt MMFs remain prevalent.
  - When MMF yields rise relative to deposit rates, investors in most regions tend to pivot away from more conservative public MMFs toward higher-yielding non-public MMFs.
  - Off-shore USD-denominated MMFs tend to grow faster compared to local currency funds, underscoring the dollar’s centrality in global cash management and demand for safe assets.

### Contextual insights
- The surge in MMF inflows during 2022–2024 coincided with aggressive global rate hikes by central banks.
- The opportunity cost of holding bank deposits increased during monetary policy tightening, as banks tended to be slower to pass on higher rates, producing a positive MMF spread and incentivizing reallocation to MMFs.
- The 2023 turmoil (including the collapse of U.S. regional banks and Credit Suisse) occurred during a globally synchronized hiking cycle, providing a unique test of substitution dynamics between bank deposits and MMFs.

### Contributions to literature
- Shifts focus from MMF outflows during systemic crises to MMF inflows during monetary policy tightening cycles.
- Provides cross-country analysis across nine countries and four regions, moving beyond U.S.-centric studies.
- Directly tests the role of the MMF yield advantage in driving MMF growth in an international setting.
- Leverages the March 2023 banking turmoil as a quasi-experiment to isolate the role of banking stress on MMF flows.
- Offers new insights on how structural characteristics—public vs non-public debt and currency denomination—shape investor behavior during rising MMF spreads.

### Implications for financial stability and monetary policy
- MMF responsiveness to rate hikes via yield differentials suggests monetary tightening can accelerate deposit outflows from banks, complicating policy transmission.
- In the absence of systemic events, MMFs appear to function primarily as instruments of yield optimization rather than as panic-driven safety switches.
- Results underscore the importance of MMF design, regulation, and transparency in managing cross-market liquidity flows in an evolving macro-financial context.

*Source: Executive Summary, "Money Market Fund Growth During Hiking Cycles" (IMF Working Paper).*

### 3. The size of the global MMF sector

### 3. The size of the global MMF sector

### MMF purpose and characteristics
- MMFs are open-ended investment funds that typically invest in diversified portfolios of short-term, safe and liquid assets.
- They aim to preserve principal value while typically offering the possibility of daily redemptions (FSB, 2024).
- In normal times, MMFs can offer comparable levels of security and accessibility, often at higher rates compared to other cash investment alternatives, such as bank deposits, and increased diversification of credit risk.

### Global scale and geographic distribution
- The global MMF sector is about 10.5 trillion euros.
- The global mutual fund sector is about 74 trillion euros.
- The MMF sector is present in a relatively limited number of countries, mainly AEs.
- The US holds the largest global MMF share, just shy of 60%.
- Europe and China account for approximately one third of the US size.
- Jurisdictions beyond the US, Europe, and China account for less than 2% of the global MMF sector size.

### MMFs as a share of mutual fund sectors
- In jurisdictions with at least 1% of the global MMF share, MMFs make up about 10 percent to 20 percent of the mutual fund sector.
- Exceptions where MMFs constitute a much larger share of the mutual fund sector include China and Mexico.
- Data correspond to 2025Q2 values.

### European MMF concentration and currency composition
- European MMFs are mainly concentrated in three countries: Ireland, France and Luxembourg.
- Ireland and Luxembourg cater primarily to a mostly non-domestic investor clientele, reflecting their roles as financial hubs.
- In Ireland and Luxembourg most funds are off-shore dollar funds, with a smaller participation of euro denominated funds and funds denominated in British pounds.
- Ireland has the largest number of British pound denominated funds.
- France hosts predominantly EUR denominated funds.
- MMF presence in other European countries is small in comparison.
- The European regulatory reform of 2017 (EU Money Market Fund Regulation, MMFR) imposed more stringent regulatory requirements and increased the cost of compliance; as a result, the number of MMFs declined notably in smaller domiciles and MMF concentration increased (EFAMA, 2020).

### Notes on figures referenced in the text
- Figure 1a: shows each jurisdiction’s share of global MMF AUM (blue bars) and the size of MMFs in each jurisdiction as a proportion of the overall mutual fund sector (yellow bars); shares computed using AUM; data correspond to 2025Q2 values. 
- Figure 1b: plots the evolution of USD- and EUR-denominated MMFs across countries over time; shares use NAV denominated in the respective currency. Data source for Figures 1a–1b is Lipper (with author calculations).

### Data: coverage, measures, and construction
- Country/sample coverage focuses on jurisdictions with large nominal MMF sectors that underwent a hiking cycle in the last couple of years: the US and Europe, as well as Australia, Brazil, Mexico and Korea; Chile is included for a more complete Americas representation. European MMFs for this analysis include Ireland, Luxembourg and France.
- Currency focus: MMFs denominated in the group of currencies that corresponds to the countries in the sample (e.g., USD funds in jurisdictions outside the US or euro denominated funds in jurisdictions outside the euro area). MMFs denominated in other currencies (such as GBP or JPY) are excluded.
- Fund-level NAV data: monthly observations from Lipper, from January 2004 to September 2023; includes fund-level information on Fund type (available only for US and European MMFs), jurisdiction and underlying currency of the Funds (available for all jurisdictions).
- Key dependent variable: year-on-year growth rate of NAVs for individual MMFs (Figure 2).
- Data cleaning: growth rates above 100 percent in absolute value are excluded to remove occasions of MMF conversions (for example from prime to government MMFs).
  - Robustness checks: results are robust when excluding growth rates larger than 75 percent and 50 percent (results not shown).
- Monetary policy variables:
  - Monthly monetary policy rates from Bloomberg used to create dummies for periods of policy rate hikes. A hiking period extends from the first hike until the month prior to the first cut.
  - A monetary policy rate shock measure is constructed as in Xiao (2020).
  - The ECB policy rate is used for euro area countries in the sample.
  - Bloomberg tickers used are listed for each country (Australia: RBATCTR Index; Brazil: Bzstseta index; Chile: Chovchov index; ECB: EUORDEPO Index; USA: FDTRFTRL Index; Mexico: Mxonbr Index; Korea: KORP7DR Index).
- MMF yield advantage (MMF spread):
  - Constructed using overnight money market rates (proxy for MMF yields) and national average bank deposit rates.
  - Overnight rates by jurisdiction are taken from Bloomberg (tickers provided in source).
  - USD (EUR) denominated MMFs outside the USA (euro area) are assigned the US (euro area) overnight rate.
  - The overnight rate may be a lower bound for MMF yields; this proxy may underestimate the MMF yield advantage for countries outside the US.
  - Bank deposit rates: data from various sources (Bloomberg, IMF, central bank statistics), focusing on savings rates or deposit rates and certificates of deposit with maturities close to one year where possible.
  - The MMF spread is defined as money market yields minus bank deposit rates, transformed as the cumulative change in the spread over the last three years; the final proxy takes both negative and positive values.
  - Observed patterns: in both sides of the Atlantic, and in Korea the spread is negative for large periods of time (bank rates offer higher yields compared to MMFs); the reverse holds in Mexico and Brazil; in Australia the difference was relatively small prior to the 2022 hiking cycle.
- Control variables: macro variables include inflation and GDP year-on-year growth rates by jurisdiction; financial variables include the VIX (CBOE Volatility Index) and yield curve spreads (10 year minus 2 years). 
  - For jurisdictions without their own yield curve spreads (other than Australia, USA and France), the US yield curve spread is assigned to EMs and dollar denominated funds, and the German yield curve spread is assigned to the remaining euro denominated funds.

*Source: Excerpt from IMF Working Paper chapter "3. The size of the global MMF sector" and "4. Data".*

### 5. Motivation

### 5. Motivation

### A. MMF yield advantage as a determinant of MMF flows
- Prior evidence: positive monetary policy shocks lead to inflows into MMFs (Xiao, 2020; Aldasoro and Doerr, 2023).
- Mechanism:
  - MMF yields adjust more swiftly during interest rate increase cycles because they are based on diversified portfolios of short-term assets whose yields closely track policy rates in highly competitive funding markets.
  - Bank deposit rates exhibit stickiness (deposit betas), influenced by market structure and bank strategies; banks with significant market power and strong “deposit franchise” in concentrated markets tend to adjust deposit rates slowly (Dreschler et al., 2017; Dreschler et al., 2021).
  - In the euro area, sluggishness in deposit rate adjustments can be due to excess liquidity in the banking system reducing the need for deposit funding (Messer and Niepmann, 2023).
- Implication: yield-oriented investors are more likely to pivot towards MMFs when monetary policy tightening widens the spread between MMF and bank rates (Xiao, 2020); conversely, monetary policy easing can disadvantage MMFs relative to banks, leading to outflows (Bua et al, 2019).
- Contribution of this paper:
  - Constructs a proxy of the MMF yield advantage—the MMF spread—and tests the hypothesis that MMFs grow faster in periods of interest rate hikes.
  - Reports that the MMF spread typically increases across countries during periods of monetary policy hikes and declines outside these periods (Figure 2); the hiking cycle of 2022 appears to have driven a strong net increase in the spread.
- Controls for broader portfolio allocation decisions:
  - Macro and financial variables used include inflation and GDP variables (also in Xiao, 2020), VIX, and the yield curve spread (10 years minus 2 years) to control for market sentiment.
  - Rationale: VIX proxies global risk sentiment; yield curve shape influences investors’ duration-taking decisions.
- Figure 3 note:
  - Charts plot average MMF spread between MMF yields (proxied as money market rates) and deposit rates across countries (blue line).
  - USD and EUR denominated funds assumed to follow USA and euro area money market rates respectively.
  - Red line denotes periods of monetary policy rate hikes. A hiking period starts at the first rate hike and ends before the first rate cut.
  - Spreads are calculated in percent.

### B. The impact of the 2023 banking turmoil
- Rationale: during banking crises investors seek deposit alternatives, which MMFs can offer.
  - Bank fragility can emerge during monetary policy tightening from interaction between declining asset values and depositor behavior shifts (Jiang et al., 2024; Copestake et al., 2023).
  - The 2023 regional banking turmoil in the US exemplified these dynamics: concerns about solvency and deposit safety triggered runs on some regional banks, leading to large deposit outflows into both systemically important U.S. banks and MMFs (Caglio et al, 2024; Adrian et al, 2024; Nikolaou, 2023; GFSR, 2023).
  - Subsequent collapse of Credit Suisse in Europe added to instability; equity prices of global banks—particularly in advanced economies—dropped sharply, while emerging markets were less affected (Pandey et al, 2023).
  - Financial stress receded quickly, aided by swift policy responses including liquidity backstops and coordinated communication by central banks (Yousaf et al., 2023; Pandey et al., 2023).
- Three features making 2023 turmoil relevant for MMF analysis:
  1. Unlike the GFC or COVID-19 shock, the 2023 turmoil did not constitute a broad systemic crisis, allowing clearer observation of substitution dynamics between banks and MMFs.
  2. The turmoil occurred amid a global hiking cycle, providing a synchronized setting to isolate the incremental impact of banking turmoil while controlling for monetary policy effects.
  3. Despite widespread narratives of sharp post-SVB inflows into U.S. MMFs and lagging European inflows, little formal cross-country analysis existed; this paper’s in-depth market analysis finds results that contradict the dominant narrative.
- Empirical focus:
  - Tests whether MMF growth during the banking turmoil was significantly higher than in previous hiking cycles, after controlling for macro-financial environment and fund-specific characteristics.
  - Investigates whether the turmoil provided an additional push to MMF inflows and how investors reacted to higher MMF spreads in that environment.
  - Seeks to distinguish whether reallocations reflected panic and risk aversion or rational shifts toward more attractive, liquid alternatives.

### C. Structure and types of MMFs as a driver of MMF flows
- MMF characteristics:
  - MMFs are open-ended investment funds investing in diversified portfolios of short-term, safe, and liquid assets; they aim to preserve principal and typically offer daily redemptions (FSB, 2021).
  - In normal times MMFs offer comparable levels of security and accessibility at higher rates compared to bank deposits, with increased diversification of credit risks.
- Heterogeneity across MMFs:
  - Broad dimension: public (government) vs private (non-government) debt MMFs—prominent in the US.
- United States:
  - Government MMFs invest in public debt and typically allow investors to get principal back on demand and at par, much like deposits; investments typically in cash and short-dated government securities.
  - Government MMFs expanded significantly following the 2015 MMF regulatory reform and currently account for more than 80 percent of the assets held by US taxable MMFs (Figure 4, Chart 1).
  - Prime MMFs are the main alternative in the US; they have broader investment options but face restrictions in redemptions.
  - The shift to Government MMFs has considerably stabilized the US MMF sector; government MMFs saw substantial inflows during the Covid pandemic, while prime counterparts witnessed outflows.
  - Government MMFs can yield less compared to non-government MMFs, which may still appeal to yield-oriented investors (Figure 4, Chart 2).
- Outside the United States:
  - Public-debt MMFs are less prominent; the relative size of public debt MMFs in the sample, excluding the US, has remained broadly stable or declined.
  - Public debt funds have a substantial share in Korea (about 40 percent) but have less than 20 percent in the remaining countries.
  - In Europe, money market funds with equivalent characteristics to US government MMFs account for only 10 percent of total EU money market funds (Constant NAVs or CNAVs).
  - The depth and size of the US Treasuries market is an important factor behind the large expansion of government MMFs and the outsized size of the US MMF sector compared to other countries.
- Private-debt MMFs in other countries:
  - Majority of MMF investments outside the US lean heavily on private liabilities, investing in diversified pools of liquid private debt that can include bank deposits (term or negotiable) in certain countries and typically offer higher returns compared to public funds.
  - These funds often allow somewhat longer maturities and higher exposure to credit risk; on-demand access to principal is typically restricted with potential fees and “gates” to safeguard liquidity.
  - Such restrictions are discretionary for US government MMFs but mandatory in Europe (Table 1).
- Currency denomination:
  - In most countries in the sample, MMFs are primarily denominated in local currency.
  - Exceptions: Ireland, Luxembourg, and, to a lesser extent, Chile. In Ireland and Luxembourg, around half of AUMs are denominated in U.S. dollars, while euro-denominated funds account for a declining share of roughly 20–30 percent. In Chile, the share of dollar-denominated funds fluctuates between 10% and 20% of the sector.
  - Overall, the U.S. dollar is the dominant currency in the global MMF market; euro-denominated funds are largely concentrated in Europe, with France hosting the largest number.
  - The data show a strong home-currency bias across markets but persistent demand for U.S. dollar–denominated assets (Maggiori, Neiman & Schreger, 2020).
- Empirical treatment in this paper:
  - Matches MMF yield to the money market rate of the currency of denomination so the spread variable aligns with both local and foreign-currency MMFs.
  - Tests whether USD-denominated funds outside the United States attract systematically higher inflows relative to domestically denominated MMFs.
- Figures and data notes:
  - Figure 4 Panel 1: evolution of public debt MMFs across countries.
  - Figure 4 Panel 2: Government vs prime MMF yields in the US.
  - Figure 5: evolution of EU Fund types.
  - Notes: government MMFs include Government and Treasury MMFs in the US and CNAVs in Europe; for other countries a fund is assigned to the category when the name suggests it is a government MMF. Data run from January 2024 to February 2024. Source is Lipper and author calculations.

*Source: IMF Working Paper — Money Market Fund Growth During Hiking Cycles (section 5: Motivation).*

### 6. Methodology and Empirical analysis

### 6. Methodology and Empirical analysis

### Empirical strategy and regression specifications
- Baseline panel regression (Eq(1a)):
  - Δy_{t,i} = β Δpr_t + controls_{t,i} + FE_{t,c,i} + ε_{t,i}
  - Δy_{t,i} denotes year-on-year growth rates of NAV at the fund level.
  - Δpr_t denotes changes in the monetary policy rate.
  - Controls: country-specific inflation and GDP, VIX, yield curve spread (10y−2y), fund-specific characteristics (country, size measured by NAV, parent company).
  - Fixed effects for time and fund characteristics included.
  - Standard errors: robust and clustered at the fund level.
- MMF yield advantage specification (Eq(1b)):
  - Δy_{t,i} = β ΔSpread_{t,c} + γ Dummy_{HC_{t,c}} + controls_{t,i} + ε_{t,i}
  - Spread is the MMF Spread proxy (cumulative difference, over the past three years, between MMF yields proxied by the overnight money market rate and the average bank deposit rate).
  - Dummy_{HC} equals 1 during a country’s monetary policy hiking cycle (from first hike until first cut).
- Difference-in-difference for yield advantage during hikes (Eq(1c)):
  - Δy_{t,i} = β (ΔSpread_{t,c} * Dummy_{HC_{t,c}}) + γ ΔSpread_{t,c} + δ Dummy_{HC_{t,c}} + controls_{t,i} + ε_{t,i}
- Post-turmoil specification for 2023 banking turmoil (Eq(2b)):
  - Δy_{t,i} = β (Period_{t,c} * ΔSpread_{t,c}) + γ Period_{t,c} + δ ΔSpread_{t,c} + ζ controls_{t,i,c} + ε_{t,i}
  - Period_{t,c} is a binary indicator equal to 1 during March–June 2023 (three months following the collapse of Silicon Valley Bank) and 0 for prior hiking cycles (adjusted to each country’s timeline).
- Fund-type specifications (Eq(3a)–(3c)):
  - Eq(3a): Δy_{t,i} = β Spread_t + γ Public_{t,i} + θ DollarFund_{t,i} + controls_{t,i} + ε_{t,i}
  - Eq(3b): add Spread * Public interaction: Δy_{t,i} = β Spread_t + γ Public_{t,i} + δ (Spread_t * Public_{t,i}) + controls_{t,i} + ε_{t,i}
  - Eq(3c): test post-turmoil preference for public funds via interaction with post-turmoil dummy.

### Impact of monetary policy on MMF growth (results from Eq(1a–1c))
- Column (1) (policy rate specification):
  - Policy rate coefficient: 4.133*** (standard error (0.125))
  - Interpretation: MMFs in sample grow on average by 2.14 percentage points for 1 percentage point increase in the policy rate (textual result); Table coefficient reported as 4.133***.
- Column (2) (Spread only):
  - Spread coefficient: 3.418*** (0.128)
- Column (3) (Spread, Dummy hiking cycle, and interaction):
  - Spread: 1.077*** (0.126)
  - Dummy (hiking cycle): 1.028*** (0.214)
  - Spread * Dummy: 0.903*** (0.155)
- Control and size effects (Table 1; all coefficients reported with standard errors):
  - Dummy (size medium): 16.708*** (0.227) in (1); 16.939*** (0.227) in (2); 15.816*** (0.284) in (3)
  - Dummy (size large): 32.058*** (0.279) in (1); 31.910*** (0.275) in (2); 31.729*** (0.347) in (3)
  - Inflation: -2.433*** (0.072) in (1); -1.028*** (0.069) in (2); -0.760*** (0.078) in (3)
  - GDP: 1.106*** (0.060) in (1); 0.751*** (0.061) in (2); 1.173*** (0.081) in (3)
  - VIX: -0.826*** (0.059) in (1); -0.437*** (0.061) in (2); -0.407*** (0.072) in (3)
  - Yield curve (YC): -2.329*** (0.071) in (1); -2.637*** (0.074) in (2); -0.388*** (0.106) in (3)
- Key empirical insights:
  - Positive relationship between monetary policy shocks and MMF growth is corroborated.
  - Larger funds grow significantly faster than smaller ones (indicating concentration).
  - MMF assets rise faster with higher GDP and decline with higher inflation.
  - MMFs experience outflows when VIX increases.
  - A steeper yield curve (rising 2s10s spread) is associated with modest outflows.

### Impact of the MMF yield advantage
- Contemporaneous correlations between yield advantage and monetary policy rates:
  - Correlation coefficients typically around 50 percent to 80 percent in most countries; exceptions: Australia and Korea (correlation coefficients almost zero).
- Table 1 (columns 2–3) summary:
  - Baseline MMF growth is higher during hiking cycles (around +1% as reported in text and represented by Dummy (hiking cycle) = 1.028***).
  - Interaction term positive and significant: effect of yield spread on flows is amplified during hiking cycles (Spread * Dummy = 0.903*** in column (3)).
  - Interpretation: investors more likely to allocate to MMFs offering superior returns when policy rates are rising.

### Impact of the 2023 banking turmoil (results from Eq(2))
- Full sample (Table 2, column (1)):
  - Dummy post turmoil: -8.873*** (1.051)
  - Spread: 3.381*** (0.172)
  - Dummy post turmoil * Spread: 2.927*** (0.561)
  - Interpretation: MMF growth appears lower in the post-turmoil period (γ negative), but sensitivity to MMF yield spread (β) increased significantly.
- Regional coefficients (Table 2):
  - USA (column (2)):
    - Dummy post turmoil: -24.984** (17.085)
    - Spread: 5.692*** (0.621)
    - Dummy post turmoil * Spread: 11.632** (5.229)
  - Euro area (column (3)):
    - Dummy post turmoil: -1.428 (2.009) [insignificant]
    - Spread: 10.045*** (0.609)
    - Dummy post turmoil * Spread: -2.272* (1.335) [negative at 10% level]
  - Americas (column (4)):
    - Dummy post turmoil: 0.044 (1.041) [insignificant]
    - Spread: 1.193*** (0.178)
    - Dummy post turmoil * Spread: 2.862*** (0.798)
  - Asia-Pacific (column (5)):
    - Dummy post turmoil: -18.709*** (5.615)
    - Spread: 2.506*** (0.502)
    - Dummy post turmoil * Spread: 5.359* (3.017)
- Key interpretations and findings:
  - Sensitivity to MMF spreads increased in most regions during the 2023 turmoil, indicating intensified yield-seeking.
  - In the U.S., post-turmoil MMF growth was lower relative to earlier hiking periods once controls are included, but responsiveness to spread rose.
  - In the euro area, growth post-turmoil remained similar to earlier hikes (γ insignificant), but spread sensitivity appeared negative at the 10% level.
  - Overall, no uniform “flight to safety” into MMFs following SVB and Credit Suisse collapses; higher MMF yields helped sustain flows post-turmoil, but overall growth was lower or similar to previous hiking periods.
  - The net effect on sensitivity to spreads remained strongly positive (β + δ > 0), suggesting normalization rather than reversal in investor behavior.
  - Headline inflows during the period largely reflect macro-financial conditions, fund characteristics, and widening yield differentials; after controlling for these, MMF inflows were smaller than during prior hiking episodes.

### Impact of MMF characteristics (fund type and currency)
- Baseline (Table 3, Eq(3a)–(3c), full sample):
  - Spread: 1.771*** (0.096) in (1); 2.051*** (0.102) in (2); 1.852*** (0.097) in (3)
  - Public dummy: 0.213 (1.005) in (1); 0.372 (1.005) in (2); 0.249 (1.003) in (3) — insignificant across specifications.
  - Spread * Public (column (2)): -1.579*** (0.185) — significant and negative.
    - Text interpretation: a 1 percentage point increase in the spread is associated with a 1.58 percentage point smaller inflow to public MMFs relative to non-government MMFs.
  - Dummy post turmoil (column (3)): -3.721*** (0.603)
  - Dummy post turmoil * Public (column (3)): -3.922** (1.195)
  - Dollar Fund dummy (offshore USD-denominated funds outside the US): 11.254*** (1.675) in (1); 11.408*** (1.675) in (2)
    - Text interpretation: offshore USD-denominated MMFs grew, on average, about 11 percent faster than those denominated in other currencies.
- Interpretations:
  - Public debt funds do not attract higher inflows relative to other MMF types in the full sample.
  - When MMF spread increases, public debt funds generally experience lower inflows compared to other fund types — indicating yield-seeking investors gravitate toward non-public MMFs (which typically offer higher yields).
  - During the post-turmoil period, public debt MMFs exhibited significantly lower growth than non-public debt MMFs.
  - Dollar-denominated funds show a strong positive growth advantage (about 11 percent faster growth).

### Regional heterogeneity in fund-type effects (Table 4, Eq(3b) by region)
- Spread coefficients by region:
  - USA (column (4)): 6.079*** (0.230)
  - Europe (column (5)): 4.913*** (0.220)
  - Europe (alternate column (6)): 10.680*** (0.301)
  - Americas (column (7)): 2.233*** (0.175)
  - Asia/Pacific (column (8)): 1.750** (0.551)
- Public dummy by region:
  - USA: 5.081*** (1.097) — U.S. public MMFs grew on average by about 5 percent more than non-public MMFs (reflects substantial expansion since 2015).
  - Europe: -3.846* (1.786) — CNAVs in Europe less attractive than non-public MMFs.
  - Americas: 4.209 (3.676) — insignificant.
  - Asia/Pacific: 10.578*** (2.832) — higher growth for government MMFs compared to other types.
- Spread * Public interaction by region:
  - USA: -4.136*** (0.281)
  - Europe: -2.702*** (0.367)
  - Europe (alternate): 1.579** (0.577)
  - Americas: -3.322* (1.438)
- Other notable regional findings:
  - LVNAVs in Europe: Dummy (LVNAV) = 5.248** (1.744) and Spread * Dummy (LVNAV) = 4.033*** (0.508) in column (6), indicating LVNAVs grew faster and attracted flows when spreads rose (LVNAVs treated as cash equivalents under EU regulation).
  - Dollar Fund effects by region:
    - USA: Dollar Fund = 11.707*** (1.829)
    - Europe: 11.333*** (1.792)
    - Europe (alternate): 11.364* (5.121)
    - Americas: 13.291*** (0.980)
- Regional interpretation summary:
  - U.S.: public MMFs (government/Treasury) grew faster on average; however, when spreads rise, public MMFs attract fewer inflows than non-public funds (negative Spread * Public).
  - Europe: CNAVs less attractive relative to other fund types; LVNAVs grew faster and attracted flows when spreads rise.
  - Asia-Pacific: government MMFs recorded higher growth.
  - Americas (excluding US): public MMFs sometimes preferred when yield advantage rises (heterogeneity noted).
  - Dollar-denominated offshore MMFs grew substantially faster across regions, reflecting preference for dollar liquidity and possibly U.S. policy rate divergence.

### Synthesis of empirical findings
- Monetary policy tightening:
  - Higher policy rates are associated with increased MMF growth.
  - The MMF yield advantage (Spread) is a significant driver of flows and its effect is amplified during hiking cycles.
- Banking turmoil (2023):
  - MMF growth overall was lower or similar to prior hiking episodes once controls are included.
  - Sensitivity to MMF spreads increased in most regions during the 2023 turmoil, indicating stronger yield-seeking behavior by investors.
  - No uniform flight to safety into MMFs following SVB and Credit Suisse; headline inflows reflect multiple factors (macro-financial conditions, fund characteristics, yield differentials).
- Fund characteristics:
  - Larger funds grew significantly faster than smaller ones (evidence of concentration).
  - Public debt MMFs do not universally attract higher inflows; when spreads rise, investors tend to shift to non-public MMFs that offer higher yields.
  - Offshore USD-denominated MMFs grew about 11 percent faster on average than non-USD funds (strong dollar preference).
- Regional heterogeneity:
  - U.S. public MMFs expanded markedly since 2015 and grew faster on average, but become relatively less preferred when spreads widen.
  - Europe saw stronger growth of LVNAVs and weaker growth for CNAVs relative to other types.
  - Asia-Pacific and some Americas jurisdictions display differing preferences reflecting market structure, investor composition, and local yield opportunities.

*Source: 6. Methodology and Empirical analysis (wpiea2025150-source-pdf).*

### Conclusion and further work

### Conclusion and further work

### Key findings
- The MMF yield advantage is identified as a primary driver of MMF growth.
- This yield advantage typically emerges during monetary policy tightening cycles.
- The yield advantage supported MMF inflows following the 2023 banking turmoil, although the 2023 banking turmoil itself did not appear to catalyze MMF flows.
- A rising MMF yield advantage can influence allocations to different fund types:
  - In the US and the euro area investors shift towards non-public debt MMFs notably when MMF spreads rise.
  - In the Americas the preference is for public MMFs.
- There is a demonstrated preference for USD-denominated MMFs in several jurisdictions outside the US, underscoring the dominant role of USD in global short-term funding markets.
- Cyclical growth behavior in MMFs can make rapid reversals of flows more likely in times of stress.

### Policy implications
- Rapid MMF growth during tightening cycles can amplify systemic risks when flows reverse sharply.
- Cyclical attractiveness of MMFs relative to alternatives (such as bank deposits) can accelerate deposit outflows, potentially complicating monetary policy transmission and straining bank funding.
- Policymakers and supervisors should ensure robust MMF regulation and oversight, including:
  - Maintaining adequate liquidity buffers.
  - Ensuring sound risk management practices.
  - Promoting transparency to help MMFs accommodate rise-and-bust cycles and mitigate systemic vulnerabilities.
- Ongoing monitoring and adaptive regulation are essential as the landscape of cash alternatives evolves, to safeguard MMF markets and the broader financial system.

*IMF WORKING PAPERS Money Market Fund Growth During Hiking Cycles — Working Paper No: WP/25/150*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2025/english/wpiea2025150-source-pdf.pdf_