Why a unilateral currency intervention is a recipe for disaster

BECAUSE the pricing of markets vary within the hour, it would be pedantic to comment on the impact of U.S. elections. As I pen this article, I think Trump is closer to win, but absentee ballots continue to pour in and they are over 90% in favour of Biden in most of the states. By Monday, when this issue of TR Monitor is released you will have the full results possibly. One thing is clear though: Either way the CBRT funding rate will continue to increase. There is not a chance to strike a bargain – I don’t know what this expression means exactly but sometime I tend to think some policy makers believe the entire market and a single shopping floor display the same mechanics, a fallacy of composition – against a nexus of market forces and fundamental dynamics that may spiral into a full-fledged balance of payment crisis. Either way I expect a steep rise in interest rates in the coming months, all of them – and also in inflation. Now let’s comment on what isn’t today – and what wasn’t yesterday – doable. What is a sure recipe for disaster in currency markets?


Can price regulations work in a fullfledged market economy? To be exact, is price regulation, i.e. controlling unwanted price fluctuations, possible in a goods (real) market or in a nominal asset market? If yes, under what conditions is that possible? Consider the currency market. Suppose you desire to keep the USD/TRY exchange rate within a band. You may declare that the initial aim is to prevent undue volatility but volatility and level go hand in hand – if markets follow any distribution one ordinarily assumes holds in asset markets. Volatility or targeted exchange rates, if reserves are to be sold anyhow, what conditions are sine qua non? Suppose a central bank believes the currency is under a speculative attack – or equivalently there is a domestic risk, i.e. dollarization. The ultimate goal ought to be to curtail the damage commodity markets (real assets or sectors) would bear as a result of nominal asset market fluctuations – in this case the exchange rate. Can we calculate from day one on the amount of currency reserves we need to use for a unilateral (quantity) intervention in such a situation? Let’s abstract from interest rate policy and its simultaneous interaction with the currency intervention, and concentrate on the mechanics of the currency (quantity) intervention alone.


Non-residents holdings of equities and bonds have been down by USD 12.850 billion over the last 12 months. So c. USD 13 billion has flown out. If this is somehow considered “hot money” or in the most impossible world of all possible worlds if some deems this a “speculative attack”, at least the amount to be compensated by the central bank should be clear from the onset. Let’s put things in perspective though.

Conjecture: The magnitude of reserves you need for a currency intervention is proportional to (a) the volume of nominal assets (funds) that are used for allegedly speculatively attacking your currency (b) to the correlation coefficients of real markets (goods markets or sectors) with each other (c) to the inverse of the exchange rate band you are targeting.

If the attack means a flow of funds that add up to USD 15 billion then the reserve you need is, say, slightly around that; if the amount doubles then he magnitude you have to sell to the market also doubles, etc. If dollarization is the risk, then selling out of reserves is again a disaster because you might need an infinite amount of reserves to counter that. Similarly, if goods markets are intricately bound up so shocks are highly correlated, then you would need an incredibly large amount of reserves. This goes back to the classic book on commodity markets by David Newberry and Joseph Stiglitz, 1981.

Finally, the band is crucial. If you try to fix the currency within a narrow band, say 8-8.25 USD/TRY, then you will need a lot more reserve compared to a larger band such as 8-9. The reason is clear: you would lose reserves trying to defend 8-8.25, and then 8.25-8.50 and 8.50-8.75, etc., without impact. On the contrary you could have a deterring effect should you intervene at one stroke at 9 only. In any case, you should be prepared to lose a good part of your reserves and you should know in advance how much you can afford to sell.

Especially given USD 185 billion FX-denominated deposits and USD 165 billion a corporate short position – and a soaring current account deficit – the required amount would be rather large. There isn’t even the trace of a speculative attack effect because non-residents only hold 4.76% of the domestic T-bill stock. Why and where have all the reserves gone? It is difficult to understand the logic behind all this because all economists should know selling reserves unilaterally would be ineffective. Now that the public has a higher borrowing requirement, rates would go up naturally, but contouring this by issuing FX- denominated domestic debt is again a sure recipe for risk if not outright disaster. Remember 2001!


Unilateral currency interventions are generally unsuccessful, and this is a parable of modern economics known for decades. If you wish to “defend the currency” then you have to act in tandem with the Fed or in any event with a supranational such as the IMF. One such period would be 1981-1985 when the dollar had appreciated to such an extent that it was on a par with the Pound Sterling. It took the Plaza Accord of 1985 to end this episode. At that time, the reason for dollar appreciation was flows to American Treasury bills because of the interest rate differential. American fixed income had become very attractive compared to the Deutsche Mark, Sterling and other major currencies in the aftermath of the Volcker deflation. Indeed, in June 1980 the FFR had hit 20%, the peak level in a century. The dollar lost value against the Deutschmark only after the multilateral intervention in 1985. The G-5 countries had called for an “orderly appreciation” of the major currencies against the dollar and the U.S. agreed to take full part in the coordinated exchange market intervention.

The impact was instantaneous. The dollar fell by 4% in a single day, and the average dollar depreciation between September 1985 and February 1986 reached 35%, USD/DM dropping from 3.47 to 2.25. Now that was indeed a successful intervention. Another example would be the Smithsonian Agreement that took place in Washington, D.C in December 1971 and ended the Bretton Woods. The G-10 had agreed to that drastic change no matter what pressure Nixon had put them under beforehand. If your currency isn’t that important in world trade, then you have little option left. One thing is certain: a currency intervention through selling out of central bank reserves to the market can only be successful if it is very well thought out and designed with cooperation of the parties involved. It has to be one-shot also. If risk is continual, as in the case of dollarization, then currency intervention is a sure recipe for eroding your precious reserves without accomplishing anything tangible.


If the currency risk goes up (a) but there isn’t one single risk factor that stands alone – suppose cross border flows, geopolitical risks, foreign trade risks, the obvious failure of your growth strategy, if any, etc (b) and the risk is endogenous – dollarization, FX-denominated domestic debt issuance, over-reliance on imports so you can produce and export, a very closely knit co-variance matrix across goods market domestically and suchlike (c) however the currency fluctuation isn’t one-shot it is continual, then the intervention will end up a failure. That much is obvious and isn’t subject to debate. It is economics at its best. In fact, even multinational efforts to regulate either currency or commodity markets with the help of the IMF or under the auspices of UNCTAD, as happened in the 1970s, didn’t end well most of the time.

Leave a Reply

Your email address will not be published.