Do High Interest Rates Defend Currencies During Speculative Attacks?

That is the question posed by a paper (pdf) by Aart Kraay of the World Bank, written in 2001 and focusing mainly on fixed exchange rate scenarios.  It seems the high rates do not protect currency values, here is the abstract:

Do high interest rates defend currencies during speculative attacks? Or do they have the perverse effect of increasing the probability of a devaluation of the currency under attack? Drawing on evidence from a large sample of speculative attacks in developed and developing economies, this paper argues that the answer to both questions is ”no”. In particular, this paper documents a striking lack of any systematic association whatsoever between interest rates and the outcome of speculative attacks. The lack of clear empirical evidence on the effects of high interest rates during speculative attacks mirrors the theoretical ambiguities on this issue.

This study (jstor) from the JPE, by Lahiri and Vegh, shows mixed results and also explains why the higher interest rates may be counterproductive.  They increase public debt service and may predict higher future inflation, thereby worsening some of the constraints and possibly hastening a further speculative attack.

Here is an Allan Drazen survey on raising interest rates to defend a currency (pdf), much of which focuses on the signaling effect.  He argues the non-signaling effects are generally weak and when it comes to the signaling effects it can cut either way.  Needing a large rate hike to defend a currency is in some ways a negative signal as well.  The empirics are discussed starting on p.51 and one result seems to be short-term benefits and medium-term deterioration, following a big interest rate hike to protect a currency.  Again, you will note this estimation is drawn from a lot of fixed rate countries and it may apply to floating rate scenarios only with qualifications.

You will find other relevant readings here.

The big news, of course, is that the Turkish central bank yesterday announced a 425bp rate hike, to stem a currency crisis.  There is FT Alphaville on the move here, more readings here.  The markets seem to like this move, at least at first, but based on what we know from the literature, we should not be too quick to think this will succeed.


So keep the short end at its equilibrium rate works, by definition, if one could compute that rate.

Very good point. It would be interesting to try to identify why sometimes it works and sometimes not. I suspect one reason is that not all heavy sell-offs are speculative attacks. If the apples are discovered to be wormy, it may be difficult to convince people not to dump them.

I'd rather read a paper on "Do Speculative Attacks Matter?"

A speculator only makes money when they cover their position. So they have to buy everything back that they shorted -- yet supposedly that doesn't affect the market as much. Why?

Who is to say that speculators ever do more to a currency than what was inevitable, given the fundamentals?


The lasting damages may come from non-speculators responding to the spurious, amplified signal generated by the speculators?

Is there any fundamental reason to expect "non-speculators" to respond to the drop more than the recovery?

Perhaps it has to do with levels of FX reserves at the central bank? For example, Argentina tries to maintain a 6.2 exchange rate to the USD, I as a speculator keep shorting ARS. The collective effect of all us speculators is to bleed the Central Bank of Argentina of all of its FX reserves, the pressure builds until eventually the CBA has to abandon their peg attempt. By the time I cover my short, the CBA is bled dry, the whole dynamic has shifted, and us speculators taking our profits is not enough to offset what the central bank did and then stopped doing. That's just a thought.

Exactly. Such attacks don't just happen. All that's happening is that the Turkish central bank has realized (far too late) that its monetary policy was far too easy and its interest rates too low to accurately reflect the riskiness of the Turkish economy.

It's the equivalent of a child trying to stuff all its toys under the bed before mom makes it upstairs to verify the room is clean. Mom is never fooled.

The rate hike by the Turkish central bank is not consistent with its higher year-end inflation forecast. Speculators are right to expect future depreciation if the central bank keeps up this charade.

Great minds think alike! I posted exactly this a few hours before TC did, on Seeking Alpha. BTW I don't know about TC but I credit the Wall Street Journal Econ blog "Real Time Economics" for bringing this paper to my attention. This is yet another example of how inflation distorts the economy and can be exploited for a short while. Whether lag due to inflation of the money supply (the basis for Keynesian fiscal 'stimulus' or IS–LM 'stimulus', a close cousin), or here, how high inflation is not to be feared in the short term, and thus can be safely ignored. It's games people play due to essentially "money illusion", as any behavioral economist will tell you.

I would say that if market participants find the rate hike to be sustainable (i.e. quasi-permanent in the short to medium term) then it acts as a defense of the currency, but if market participants instead think it is unsustainable then it is more like waving a red flag at a bull and it will add fuel to the fire (e.g. "oh no, the currency is going down; everybody get your money out now while you can") and make the currency collapse come faster and perhaps deeper than otherwise would have been the case.

My brother suggested I might like this blog. He was entirely right.
This post actually made my day. You cann't imagine just how much time I had spent for this info!

It seemed to work when the DKK was attacked back in the early nineties. The SKK was rightfully attacked (and ended up being devalued) but the other Krone currencies were attacked, too, for no good reason. The DKK kept its anchor to the DM and the interest rates were reduced again.

Comments for this post are closed