Higher interest rates – stronger currency?

In the comments Kimble had a great point “long term estimates of the level of the dollar can be heavily influenced by the current level”. Although this may seem like economists being myopic, it has more to do with our extremely limited understanding of how the exchange rate works.

When doing ‘applied’ (adhoc) work, me and economists like me enjoy using the assumption that higher interest rates will lead to an appreciation in the currency. The fundamental assumption driving this is that if interest rates are higher, they will attract more foreign capital, increasing demand for the local currency, and thereby driving up its value. Backing up this idea is a second factor, that if interest rates are higher domestic consumption falls, reducing the demand for imports at a given exchange rate, which reduces the supply of currency, increasing its value.

Now for some reason, the evidence does not follow this well polished argument. In fact, economists agree that a random walk seems to outperform any exchange rate model we can create.

What could be the problem with our description? Firstly we have expectations. Once market participants have a view of where the exchange rate may be, they might hold it there even if interest rate differentials change. I don’t believe this, as I think market participants would try to take advantage of any immediate arbitrage opportunity that presented itself. However, this can be used to explain why exchange rates can get ‘stuck’ at certain levels.

This leaves us with the sad realisation that there is probably something wrong with the model. I have no problem with the supply view, higher interest rates will lower domestic demand. However, I do have a problem with the demand section.

We are told that higher interest rates will attract capital, as it increases the return for investors. Now a higher interest rate will mean that someone can put money in the bank and make a greater return, but doesn’t it also impact on the growth in the value of assets in a negative way? In NZ at the moment, currency fluctuations seem to be driven by European and Japanese investors who simply want to put their money in our banks – in this case a higher interest rate has increased the capital inflow, which in turn has led to an appreciation in our dollar.

However, bonds and bills are one type of investment good. There can also been investment in property and the stock market. Higher interest rates will reduce peoples willingness to borrow and over the following months they will lead to a reduction in domestic economic activity. If this leads to a slowdown in the rate of property price growth (or a reduction on rental yield growth), property becomes a less attractive investment. The same holds true for stocks. If current global investors are moving into property and stocks, then an increase in the interest rate will actually reduce capital inflows, and possibly lead to a depreciation in the exchange rate. To find out how you can invest in the right places and exploit the market in real time see MBoxWave.com.

As a result, it is important to look at what investments foreign capital will move to and from before we can state whether a higher interest rate will lead to a higher exchange rate. As we can’t seem to forecast the intricacies of these individual movements, our exchange rate forecasts become very path dependent (for a similar example look at forecasts of oil prices 🙂 )

Add to FacebookAdd to DiggAdd to Del.icio.usAdd to StumbleuponAdd to RedditAdd to BlinklistAdd to Ma.gnoliaAdd to TechnoratiAdd to FurlAdd to Newsvine

15 replies
  1. CPW
    CPW says:

    I can’t agree with your “ad hoc” assumption, surely a lot of economists stick with the work-horse model (uncovered interest parity) that currencies with high interest rates depreciate. If you don’t accept that on some level, investing heavily in carry-trade strategies is a winning strategy – that is what “random walk” implies.

    The most likely reconciliation of the NZ currency/interest rate story probably starts not with the fact that are interest rates are high, but that most of the increases were unexpected (relative to earlier future curves). That would get you the observed result of rising interest rates and a rising Kiwi, but still with the expectation that the Kiwi would depreciate in the future.

    Having said that, that random walk result is supposed to hold even over forecasts using ex-post data.

    If you believe “random walk” Matt, why go looking for a more complicated story? I don’t quite follow the causation you’re suggesting in the rest of your post, its not clear to me whether you’re ascribing higher interest rates to a demand or a supply shift.

  2. Matt Nolan
    Matt Nolan says:

    I agree that economists generally use interest rate parity, but that doesn’t stop it being an empirically weak model. Also UIP is untestable, it is useful to explain what has happened, but not necessarily to predict what is going to happen.

    My fundamental point was that an increase in interest rates (ceteris paribus) does not necessarily imply that there is going to be a greater return to investors. As a greater return to investors is required to explain a rightward shift in the demand for currency the argument that higher interest rates cause a higher exchange rate is not strictly true.

  3. CPW
    CPW says:

    UIP seems fairly testable to me, it just normally fails the test. It still generates a prediction for what is going to happen, and the implication is that if that forecast is persistently biased then there are $100 bills lying on pavement waiting for people to pick them up. So either UIP is right, or people are stupid, which always suggests to me a fair chance that someday soon UIP will start being right.

    How does your argument differ from saying UIP holds, in which case investment returns denominated in the foreign investor’s currency are independent of domestic interest rates? Otherwise by definition higher (real) interest rates must result in a higher returns, no?

  4. Matt Nolan
    Matt Nolan says:

    I’m not sure how testable UIP is given that it relies on ex-ante expectations.

    I agree that my argument isn’t shutting out UIP, but I’m not sure I’m saying that the return to investors is independent of the interest rate. I’m just trying to show that the relationship between the interest rate and investor returns could be positive or negative.

    I think my argument got muddled by trying to put down that point by Kimble and attacking exchange rate forecasts. I only did it because it felt so good 🙂 . The initial reason for writing this was just to say “Hey, a lift in the OCR may not lead to the NZ dollar going up”.

  5. William Taylor
    William Taylor says:

    I think what matt is trying to get at is that if interest rates go up this will reduce risk premiums making risky assets in that country less attractive. Bonds aren’t the only thing that people invest in, if the risk adjusted returns on NZ stocks were to fall that would cause people with money invested in stocks to pull money out of the country. It would also cause people who invest in bonds to put money into the country. So it isn’t just a one way street.

    the main thing here is that money gets moved between countries for more reasons then chasing interest rates, UIP is an arbitrage condition related to interest rates, PPP is an arbitrage condition related to goods. UIP also ignores risky assets which for a typical investor will probably make up around 50%(depends heavily on risk aversion) of their portfolio.

    So I have to agree with your statement

    “So either UIP is right, or people are stupid, which always suggests to me a fair chance that someday soon UIP will start being right.”

    UIP makes sense, it just doesn’t tell the whole story

  6. Matt Nolan
    Matt Nolan says:

    “will reduce risk premiums making risky assets in that country less attractive”

    That was sort of what I was saying. Reducing the risk premium will reduce demand for that asset relative to bills and bonds, and so in that sense we should experience a shift from one type of investment to another. As a result, in this case I think a higher interest rate would increase demand for currency.

    The main thing is that higher interest rates need to reduce the yield associated with a given asset for my attempted line of reasoning to work. If higher interest rates only reduce the price of assets, but then don’t reduce the yield associated with the asset we should see a greater quantity of assets being bought (although this depends on the price elasticity of demand for an asset). As a result, if higher interest rates significantly reduce the growth in rental income associated with property, or lower the dividend associated with stocks (or the expected growth in the price of either of these assets), then there may be a case for a ceteris paribus jump in interest rates to lower the value of the currency.

    Now both of you know a lot more about this than me, I do realise that I’m just vomiting out my uninformed opinion 🙂 . I don’t think we have significantly disagreed though, but that might be because I don’t know what I’m talking about 😉

    About UIP though, I’m positive its empirically unfalsifiable in its most general form, as you have to make an assumption about expectations to test it.

  7. CPW
    CPW says:

    I’m not completely familiar with the evidence but I was pretty certain that UIP hasn’t tended to hold either using ex-ante forecasts for interest rates from futures curves, or assuming perfect forecasts and using ex-post interest rates. Now obviously that doesn’t rule out that idea that investors had some other expectations not closely modelled by ex-ante futures or ex-post rates but otherwise that seems about as falsified as you’re likely to get in economics. Although of course there are plenty of efforts to resuscitate the model with varying risk premiums, home country bias etc.

    As to agreement, i totally agree with the conclusion that higher interest rates need not lead to appreciation – that’s what the empirical evidence says.

    But I’m still not happy with the explanation. A change in domestic interest rates doesn’t change the risk premium for domestic equities versus foreign rates or foreign equities (assuming no change in domestic share prices) , so I don’t see the currency effect occurring as described. If instead we using a standard model for share prices an increase in the discount rate would just produce an instantaneous drop in the price, but afterwards expected returns would be higher as well.

    I agree that the effect your describing is possible assuming that tighter monetary policy leads to either expectations of reduced future company earnings or higher future volatility. The first seems analogous to the argument made sometimes that a increase in short-term rates can lead to a reduction in long-term rates (if investors believe that the Reserve Bank has tightened too much and increased the risk of recession, say). The relative magnitude of the effects still seems all wrong to me though, short-term interest rates should have more of an effect on cash allocation decisions that equity allocation, and foreign holdings of NZ assets are much more concentrated in interest bearing assets than in equities I think.

  8. William Taylor
    William Taylor says:

    “Using a standard model for share prices an increase in the discount rate would just produce an instantaneous drop in the price”

    By standard I assume you mean the CAPM right? In the CAPM the effect of the risk free rate on the discount rate depends on the beta of the company involved. Since we are talking about the market as whole we can think about beta=1, in which case an increase in the interest rate will not change the discount rate/expected return.

    You’re probably very right that foreign holdings of new zealand assets is focused on interest bearing assets.

    However foreigners aren’t the only people who move money in and out of new zealand though. if the risk/return profile of new zealand assets worsens domestic investors might shift towards overseas assets which would result in an increase in the supply of the new zealand dollar, depreciating the exchange rate.

    I just don’t think the real story is as simple as UIP makes out since it generally ignores risk and the goods market.

    By the way I’m not trying to defend the CAPM, it probably has more problems then UIP!

  9. CPW
    CPW says:

    I was thinking simpler sorry ie future dividends discounted by real rates. But even in CAPM an increase in the real rate results in a efficient market portfolio with a higher return than before doesn’t it?

    Just thinking it through, if dividends are held constant, and an increase in interest rates leads to domestic investors desiring a shift from shares to bonds, then share prices fall implying that expected returns are higher. What am I missing?

  10. Matt Nolan
    Matt Nolan says:

    Chris, I agree, thats why I said:

    “if higher interest rates significantly reduce the growth in rental income associated with property, or lower the dividend associated with stocks (or the expected growth in the price of either of these assets), then there may be a case for a ceteris paribus jump in interest rates to lower the value of the currency”

  11. Kiwi Trader
    Kiwi Trader says:

    I threw all the economic theory out the window long ago. It hardly ever worked, and when it did it was not repeatable.

    I have now been trading markets for over 30 years, and am still around making money.

    In looking at an exchange rate there are about 10-15 reasons why they go up and down. The problem is that markets place greater or lessor weight depending on how they feel at the time. Money supply in the US was a big deal once, now it is irrelevant. US housing is now the big driver, but who cared two years ago.

    So the trick is to figure what is important at the time. Also, in any exchange rate, there are two sides to the debate. In the case of the NZD/USD there are two economic drivers, and they take turns in being important. Some days it is US data, other days NZ data.

    My conclusion is that once you have figured out what is going on, it isn’t going on anymore, as the rest of the world has worked it out too. You have to take into account that the world has changed once they work it out and compensate.
    You always have to be ahead of current writing, as that is always reporting what has/is happening, not the next move.

    Equaly I have no faith in any economic forecasts beyond three months. They are almost always wrong. If a bank economist could regularly get it right, they would not work there.

Trackbacks & Pingbacks

  1. […] Read the rest of this great post here […]

  2. […] our target for some reason, we increase interest rates, this also increases the exchange rate (not necessarily the case, but seems to happen anyways). Very […]

Comments are closed.