Executive Summary

Central-bank decisions are among the most closely watched events in global financial markets. For emerging-market currencies, changes in developed-market interest rates can alter capital flows, funding conditions, risk appetite and the relative attractiveness of domestic assets.

This study asks a simple question:

Do larger G10 policy-rate changes systematically lead to higher emerging-market currency volatility?

Using 206 effective G10 policy-rate changes between November 2016 and July 2026, I examine five-day realised volatility across six emerging-market currencies: the Brazilian real, Mexican peso, South African rand, Indian rupee, South Korean won and Thai baht.

The results are less straightforward than the conventional intuition might suggest.

Across both the aggregate EM FX basket and the six individual currencies, the magnitude of the observed policy-rate change does not robustly predict subsequent five-day currency volatility.

Instead, broader global financial conditions appear more informative. In the full sample, VIX has strong positive relationships with volatility in BRL, MXN and KRW, even after correcting for multiple hypothesis testing.

However, several of these relationships weaken substantially when the extreme COVID-era period is removed.

The evidence therefore suggests that:

The market environment surrounding a policy decision may matter more for EM currency volatility than the absolute size of the observed rate change itself.

A key limitation is that the analysis uses actual policy-rate changes rather than the unexpected component of monetary-policy decisions. Markets price expectations in advance, meaning that a widely anticipated 25bp increase may contain far less information than an unexpected decision.

1. Research Question

The central research question is:

How are G10 policy-rate changes related to short-horizon volatility in emerging-market currencies?

Four initial hypotheses were considered:

Larger G10 rate changes would be associated with higher EM FX volatility.
Federal Reserve decisions would have a stronger relationship with EM FX volatility than equivalent changes by other G10 central banks.
The relationship between policy changes and EM FX volatility would strengthen during periods of elevated global risk aversion.
Sensitivity would differ substantially across individual emerging-market currencies.

2. Data

G10 monetary policy

Policy-rate data are drawn from the Bank for International Settlements Central Bank Policy Rates dataset.

The analysis covers:

Federal Reserve
European Central Bank
Bank of England
Bank of Japan
Swiss National Bank
Bank of Canada
Reserve Bank of Australia
Reserve Bank of New Zealand
Riksbank
Norges Bank

An event is identified when the daily BIS policy-rate series changes.

The final sample contains:

Emerging-market currencies

Daily exchange-rate data come from Federal Reserve H.10 series distributed through FRED.

The six currencies are:

BRL — Brazilian real
MXN — Mexican peso
ZAR — South African rand
INR — Indian rupee
KRW — South Korean won
THB — Thai baht

For each policy event, I calculate annualised realised volatility using the first five available trading-day returns following the event.

A 20-trading-day pre-event volatility measure is also constructed.

Global financial controls

The regressions incorporate:

VIX
broad trade-weighted US dollar index
US 10-year Treasury yield
pre-event EM FX volatility
hike/cut indicator
Federal Reserve event indicator

3. Methodology

The core empirical framework estimates post-event EM FX realised volatility as a function of policy-rate changes and global financial conditions.

A representative specification is:

\[ Volatility_{t+5} = \alpha + \beta_1 RateChange + \beta_2 Fed + \beta_3 VIX + \beta_4 RateChange \times Fed + \beta_5 RateChange \times VIX + Controls + \epsilon \]

Policy-rate magnitude and VIX are mean-centred before constructing interaction terms.

Heteroskedasticity-robust HC1 standard errors are used.

Separate regressions are also estimated for each of the six emerging-market currencies.

Because multiple currency-level hypotheses are tested simultaneously, Benjamini-Hochberg false-discovery-rate corrections are applied.

4. Bigger Rate Moves Do Not Consistently Produce Higher EM FX Volatility

The raw scatter shows only a weak positive relationship between the absolute size of a G10 rate change and subsequent EM FX volatility.

More importantly, once broader financial conditions are controlled for, the policy-rate magnitude coefficient is not statistically significant.

This pattern also persists across the six individual currency regressions.

The lack of a robust relationship is economically important because it suggests that the numerical size of a central-bank move is not necessarily equivalent to the amount of new information entering financial markets.

A 25bp increase that was completely anticipated may create less market disruption than an unexpected hold, cut or change in forward guidance.

5. Global Risk Conditions Matter More in the Full Sample

The full-sample relationship between VIX and EM FX volatility is considerably stronger visually.

In the aggregate model, higher VIX is associated with greater post-event EM FX volatility.

However, the extreme observations during periods of acute market stress immediately raise an important robustness question:

Is this a stable relationship, or is it being driven by crisis episodes?

6. Currency Sensitivity Is Highly Uneven

The six currencies respond very differently to global risk conditions.

In the full sample, the strongest positive VIX coefficients are observed for:

After applying the Benjamini-Hochberg correction across all tested currency-coefficient combinations, BRL, KRW and MXN retain statistically significant VIX relationships.

This heterogeneity is itself informative.

Emerging markets should not be treated as one homogeneous asset class. Currency behaviour can reflect differences in:

external financing requirements;
commodity exposure;
domestic interest-rate structures;
reserve adequacy;
trade relationships;
investor positioning;
liquidity;
sensitivity to global dollar funding conditions.

7. Crisis Conditions Matter

The most important robustness exercise removes the period from March 2020 through June 2021.

The chart shows that several full-sample VIX relationships weaken materially when this extreme market period is excluded.

The aggregate interaction between rate-change magnitude and VIX also loses statistical significance outside the COVID period.

After multiple-testing correction, none of the individual currency-level effects in the no-COVID sample remains statistically significant at the conventional 5% level.

This materially changes the interpretation.

The evidence does not support the claim that VIX has an equally strong and stable relationship with EM FX volatility across normal and crisis conditions.

Instead, periods of extreme global stress appear to account for a meaningful share of the full-sample relationship.

8. The Federal Reserve Result

Another initial hypothesis was that Federal Reserve decisions would have a disproportionately large relationship with emerging-market currency volatility.

The evidence provides little support for this in terms of rate-change magnitude.

Across the aggregate model and individual currencies, the interaction between:

Rate-change magnitude × Federal Reserve event

is generally statistically insignificant.

This does not mean the Federal Reserve is unimportant to emerging markets.

Rather, it suggests that the absolute numerical size of the Fed’s observed policy change is not, by itself, a reliable measure of the information shock reaching markets.

That distinction is crucial.

9. Why Monetary-Policy Surprises May Matter More

The main limitation of this project also provides the most natural extension.

Suppose markets expect:

Federal Reserve: +25bp

and the Federal Reserve delivers:

+25bp

The dataset records a 25bp policy change.

But the surprise to investors may effectively be close to zero.

Now suppose markets expect +25bp and the Federal Reserve instead delivers +50bp.

Both cases involve monetary tightening, but the information content of the second decision is much larger.

Financial markets therefore respond not only to:

What did the central bank do?

but also:

What did it do relative to what investors had already priced?

A stronger next-stage model would therefore replace raw rate changes with monetary-policy surprises measured around announcement windows.

This could potentially distinguish:

target-rate surprises;
forward-guidance surprises;
balance-sheet or quantitative-easing surprises.

10. Model Fit Across Individual Currencies

The explanatory power of the macro variables differs considerably by currency.

This reinforces the broader conclusion that emerging-market FX responses are heterogeneous rather than uniform.

11. Limitations

Several limitations should be emphasised.

First, policy changes are not policy surprises.
The model does not directly observe market expectations immediately before each decision.

Second, the BIS series records effective policy-rate changes.
The effective date is not necessarily identical to the precise announcement timestamp.

Third, monetary policy is endogenous.
Central banks change rates in response to economic conditions that can themselves affect currency volatility.

Fourth, crisis periods are influential.
The COVID robustness exercise demonstrates that extreme observations materially affect some relationships.

Fifth, the event window is deliberately short.
Five-day realised volatility captures immediate market behaviour but may miss slower transmission channels.

Sixth, the sample covers six EM currencies.
The results should not automatically be generalised to the entire emerging-market universe.

12. Conclusion

This project began with an intuitive hypothesis: larger G10 monetary-policy changes should generate greater volatility across emerging-market currencies.

The data provide little robust evidence for that proposition.

Across 206 G10 policy-rate changes between 2016 and 2026, the magnitude of the observed rate move does not consistently predict five-day EM FX volatility, either at the aggregate level or across six individual currencies.

Broader global risk conditions show stronger relationships in the full sample, particularly for BRL, MXN and KRW. Yet several effects weaken materially once the extreme COVID period is excluded.

The central conclusion is therefore more nuanced:

For emerging-market currencies, the environment surrounding a central-bank decision appears more informative than the absolute size of the observed rate change itself.

For investors, this suggests that analysing monetary policy requires more than tracking whether a central bank moved rates by 25bp or 50bp.

The more important questions may be:

What had markets already priced?
How surprising was the decision?
What is happening to global risk appetite?
What is happening to dollar funding conditions?
Which currencies are most exposed to those conditions?

The natural next stage of the research is therefore to move from policy changes to policy surprises.

G10 Monetary Policy and Emerging-Market FX Volatility

What 206 Policy-Rate Changes Suggest About Risk, the Dollar and Monetary Policy

Oluwadamola Shitta-Bey