Brazil and Chile on FX Intervention Watch

The Brazilian (BCB) and Chilean (BCCh) central banks have fulfilled their immediate goals in the FX markets last year, but the outbreak of the coronavirus has put us back on intervention watch. In short, we think that the bar is lower for the BCB to re-start intervening than for the BCCh. The recent sell-off, along with weaker economic data in Brazil, has brought USD/BRL back into the perceived intervention zone, based on the BCB’s recent actions. In Chile, the BCCh’s $20 bn “bazooka” will function as a free “put” against the risk of more social unrest or other downside tail risks, but aside from that, we expect no more actions in the near term.


The BCCh has fallen back in its traditional hands-off approach, but officials made it clear that they won’t hesitate to act if the price action becomes extreme. The peso has been one of the worst performing currencies year to date, yet the BCCh halted the intervention program initiated last year. In other words: they are done, and now it’s up to markets to clear CLP. That said, the program is still in place and the $20 bn “bazooka” is armed, serving as a free “put” protecting investors against downside risks in case social tensions (or other tail risk) flair up again. We don’t have a strong directional view for USD/CLP at this point, expecting it to revert to a moderate appreciation trend along with most EM currencies once the dust settles. But we also think it offers a good asymmetrical risk for being long: the BCCh’s protection against downside tail-risks is available now, and nobody knows for how long it will remain active.

Here is a recap of the last major intervention programs by the BCCh. Note that BCCh has a distinct and consistent style of intervention, announcing large programs with specified amounts and end dates, even if they don’t carry them through to the end. Still, the BCCh couldn’t be any more different to central banks in EM Asia, for example, which tend act more covertly and with scarce communication.

  • August 2001 (Selling USD): Pre-announced sales of USD spot and an increase in supply of USD-Denominated Bonds (BCD) by up to $2 bn for each, with no time specification. In the end, the BCCh sold less spot and more BCD than planned. Context: CLP under pressure from Argentina crisis and 9/11 attacks.
  • October 2002 (Selling USD): Pre-announced sales of USD spot and an increase in supply of BCD by up to $2 bn for each and set the program’s duration to 4 months. But it decided not to carry through the spot part of the program and fell short on its BCD target. Context: Financial stress in the region (Argentina, Brazil, and Uruguay).
  • April 2008 (Buying USD): Pre-announced purchases of $8 bn though daily purchases. The program fell short of the purchase target by 30% due to the collapse of Lehman (see below). Context: CLP appreciation leaving the “real exchange rate […] below the level that should prevail once real and financial conditions in the world economy return to normal.” (BCCh statement).
  • Sep 2008 (Selling USD): Offering USD via swaps to local institutions. Mostly a liquidity provision given the global shortage of USD funding. Context: Global Financial Crisis.
  • Jan 2011 (Buying USD): Pre-announced purchases of $12 bn spot though daily purchases. Context: CLP appreciation and “high commodity prices, low interest rates, slow recovery of developed economies, and depreciation of the US dollar.” (BCCh statement).
  • Nov 2019 (selling USD): Pre-announced sales of up to $10 bn in spot and $10 bn in FX hedging instruments. The bank continued rolling over forwards into 2020 but stopped selling spot. Context: nationwide protests.

(For references and a good discussion of BCCh intervention, see this IMF post.)


Intervention by the BCB is far more frequent than by the BCCh, almost dependable, and very focused on alleviating liquidity strains. For that, the BCB often uses instruments that (in theory) don’t impact the level of the exchange rate. For example, through the second half of last year the BCB was providing dollars in the spot market matched by equivalent drains via swaps (bottom panel, “conjugado”). This was designed to meet higher demand for spot as corporates pre-paid foreign-currency debt and switch to BRL debt. Before that, the BCB had been active in providing USD spot credit lines as temporary liquidity injections (top panel of the chart below). While these auctions shouldn’t impact the currency level, they might serve as a signing mechanism to show that BCB is watching and present in the market.

What about directional interventions? Since mid-2013, the BCB has been constantly intervening through FX swaps, leaning against broad trends. A cursory look at the ebb and flow of the BCB’s swap book is enough to see how they have used their FX swaps both during periods of prolonged one-way moves (appreciation and depreciation), and to counter episodic excessively volatile or “dysfunctional” markets—arguably like the one we are going through now.

Does the level of the real matter? We won’t pretend to know the answer to this hotly debated question. But looking at the charts, it’s hard to escape the inference that the BCB’s criteria for excess volatility often gets triggered when USD/BRL is around the R$4.20 level: the last two spot interventions (August and November 2019), and the swaps in August 2018 and in August 2015. That said, there were several instances of directional intervention well below the R$4.00 level, and a few times that USD/BRL made a swift move above R$4.00 and there was no intervention. If aliens arrive in Sao Paulo to study the Brazilian FX markets, with no understanding of political context or earthly central bank communication, we think they might conclude something like this: a fast rise in USD/BRL towards the R4.20 level is not a necessary or sufficient condition for directional FX intervention by the BCB, but it’s a very strong indicator that intervention is likely.

The BCB could try to use broad EM FX trends as an alibi against accusations that if defends a level, but this alibi has been looking less credible after the last few episodes. As a proxy, we take JPM’s EM FX index and Bloomberg’s LatAm FX Index, both as a ratio of BRL (FXJPEMCS/BRLUSD and LACI/BRLUSD). This gives us a simple way to measure periods of BRL underperformance and outperformance against a basket of other EM currencies. On the basis of these measures (admittedly arbitrary), it would be hard for BCB to justify some of their actions. For example, the first spot sale on 27 of August 2018 seems inconsistent with previous ones: BRL was underperforming somewhat against other EM currencies, but less than in recent episodes when there was no intervention. And if relative performance against EM really were an important criterion, the BCB should have probably intervened this year. In any case, the current moves will prove a good test to falsify or not these conjectures.


Final Thoughts

Despite the differences, we hold both the BCB and BCCh in the highest regard in terms of their FX intervention policy. We think they are effective and creative in using instruments to fulfil their objectives. Moreover, both countries have enough reserves to credibly back up any intervention program. In terms of communication, transparency, and availability of data, the FX intervention programs of both central banks are miles ahead of their counterparts in Turkey and many EM Asia countries (which we plan to write about eventually).