Tuesday, June 7, 2011

An Advanced Discussion on Krugman's Disastrous Totally Screwed Up Thinking (Wonkish)

Following my post, Krugman's Disastrous Totally Screwed Up Thinking, Bob Murphy emailed me to get me to clarify my position that Krugman is wrong about his position on real rates.  Following a back and forth of emails with Murphy, I provide below a more detailed, refined explanation of why Krugman is really out of his mind to think that the TIPS base rate reflects the real rate.

For starters, most economists consider the "real" interest rate as the nominal rate minus the inflation rate. (I would define it a little differently, but I can roll with this for the sake of argument, so as not to bring in other extraneous factors not at the essence of Krugman's confusion)TIPS pays out  the inflation rate (measured by the CPI) on top of a fixed rate. TIPS call this rate the "real" rate, but as we shall see, I believe this should be called the "base" rate to eliminate confusion. A confusion that has trapped Krugman into a pretty nutty theory, namely that the TIPS base rate is the real interest rate. Indeed at one particular point, Krugman says the "real" rate on Five year TIPS is negative. When it is in fact the base rate and not what economists would consider the real rate.

Here's is why Krugman is confuse. Comparing the real rate to the TIPS base rate is like comparing apples to apple juice. There are some similar attributes, but I wouldn't recommend drinking an apple or biting into a glass of apple juice, and that's what Krugman does.

If we take the mainstream definition of the real rate, then on a five-year fixed rate Treasury security you have a nominal rate and a real rate (The nominal rate minus the inflation rate equals the real rate) I  TIPS is different. There are other factors that play into the yield.

With the five-year fixed Treasury security you have all kinds of bidders, some who think there will be high price inflation in the future, some who think there will be deflation (such as Krugman has believed at certain times) and others who aren't really thinking about inflation either way. Thus, the pool of bidders is a mixed pool with varying views about inflation.

This pool of bidders ultimately results in market price for the five year Treasury. But, for all practical purposes, it's pretty clear that five-year Treasury buyers are all pretty much price takers. Here's the rate, take it or leave. So there is a very real possibility that someone concerned about inflation may see the yield for fixed yield securities not offering enough inflation protection and thus may want to buy added protection via a call option on an inflation hedge, such as gold, the CRB index, etc. 

It is this subset of all five-year Treasury security buyers, that are going to be most interested in buying TIPS, i.e.,those concerned about future inflation. For them, TIPS are an alternative to buying the five year Treasury security plus a  call option. If this subset thinks price inflation is going to be stronger than 1.6% (the current five year fixed rate), they will have two choices, either hedge their fixed Treasury purchase by buying some type of call option on inflation or alternatively buy TIPS, as long as the cost of buying does not result in a lower base rate than 1.6% earnings over the five years less the cost of the call option on inflation. So there is more to this TIPS "real" (base) rate than just the nominal rate minus the inflation rate. There is a built in call option price that investors will factor in when buying TIPS. This right here totally obliterates Krugman's argument that "Thanks to inflation-protected securities, we can look at real rates directly." He has simply failed to take into account that the TIPS base rate includes the cost of a call option on inflation, and thus he is way off the path when he thinks a negative TIPS means a negative real rate.

By Krugman looking at the TIPS buyers and coming up with a real rate, he is failing to consider the costs of call options, which is very likely to be a built in factor.

But that is only the start of Krugman's confusion.  As I write in my post:
Alternatively, if for some reason I expect the TIPS inflation adjustment to come in much higher than real inflation and alternative interest rates, I would be willing to buy a negative current yield for a TIP security.

The point is that there are a number of reasons a TIP security can go negative, none having anything to do with the real interest rate.
Think of it this way. Say we know that price inflation is a lagging indicator, and we know one more reading in the CPI is going to come in at 10%, but the stock market just crashed the week before the CPI is due out, and thus expectations across the board are that inflation is not on the horizon, yet the CPI number coming out will be higher because it measures a period before the stock market crash. In other words it will pay off for at leats one period on TIPS after the crash at an annual rate yield of 10%

Thus, the possibility does exist for something like this. A five year fixed Treasury might crash from a current yield of 10% to, say, 3%, because no one expects inflation after a period any more. This would likely cause the TIPS to go, if there is a base TIPS rate of 4%, to at least negative 11% as it adjusts for the upcoming CPI payment(s), even though no one in the world believes that the real rate is negative 11%. (The actual computation gets even more complex and depends on how much time exists before the five-year TIPS matures and how many actual high-yield CPI numbers come in, but the general concept is what I am trying to get across here rather than a specific negative yield number) This negative base number would have nothing to do with the real rate and also have zero to do with a major inflation expectations. In fact it means, under this scenario the opposite of an inflation expectation. While at other times, a negative base rate could reflect high inflation expectations and an adjustment for a high cost of a call option on inflation.

Bottom line the TIPS instrument is very complex and Krugman is completely confused  in thinking that "Thanks to inflation-protected securities, we can look at real rates directly". TIPS do no such thing.


  1. This is fixed income 101. Krugman is simply an academic with little (to no?) trading market knowledge.

  2. But is TIPS really a treasury + call option? Because during deflationary times the latter would be worth more than TIPS

  3. Wenzel calls it a "call option on inflation" and uses other markets such as gold and the CRB idx as examples. But for Treasurys, yes it would be a put option on the Treasury security itself. Incidentally, the Fed might be massively short just this very product:


    A final aside: the seasonality and volatility of the CPI time series itself account for much of the variation in TIPS prices (and inversely their yields).

  4. Measuring the difference between TIPS and nominal treasury rates gives you the 'market based expectation of future CPI' over the duration of the security, i.e., it is a measure of 'inflation expectations' (I'm using the scare quotes because inflation to my mind is an increase in the money supply, not the rate of change of a price index). Thinking about 'real rates', they are probably best defined as the difference between nominal rates and market-determined inflation expectations. So there is some information in TIPS yields, but it is indirect. There will never be a reliable way to 'measure' the decline in money's purchasing power, so the determination of real interest rates will never be more than an approximation.