Here is an excerpt from a recent TSI commentary.
Despite the popularity of doing so, subtracting the percentage change in the CPI or some other price index from the current nominal interest rate will not result in a realistic or reasonable estimate of the current ‘real’ interest rate.
The method of real interest rate calculation summarised above is wrong in two different ways, each of which is sufficient to render the result invalid. The first and most obvious way it is wrong is that the CPI does not reflect the change in the purchasing power of money. This is not just because it has been re-jigged over the decades as part of an effort to minimise its value, but also because the entire concept of a “general price level” is nonsense. There is no such thing as a general price level because disparate items cannot be averaged. To explain by way of a simple example, averaging the prices of a car, a potato and a visit to the dentist makes no more sense than averaging the goods/services themselves. Clearly, a car, a potato and a visit to the dentist cannot be averaged.
However, even if, for the sake of argument, we assume that the CPI makes sense at a conceptual level and is a satisfactory estimate of the change in the purchasing power of money, we still couldn’t use it to determine the current real interest rate. The reason is that the real rate of return obtained from an interest-producing investment has nothing to do with the historical change in the purchasing power of money and everything to do with the amount by which the purchasing power of money will change in the future. For example, if you buy a 1-year bond today your real return will be determined by how much the purchasing power of money changes over the next 12 months; not by how much it changed over the previous 12 months.
So, when you see a chart showing the nominal interest rate minus the 12-month percentage change in the CPI, what you are looking at is NOT a chart of the real interest rate.
How, then, should the real interest rate be calculated and charted?
The hard reality is that there are some things worth measuring that simply can’t be measured. The real interest rate falls into this category. By taking into account money-supply growth and population growth and by making a guess regarding productivity growth it is possible to come up with a realistic, albeit very rough, estimate of how the purchasing power of money shifted over a long historical period, but it will never be possible to calculate the current real interest rate.
The best we can do is use the financial market’s average forecast regarding the future CPI in our calculations. In other words, the best we can do is use the TIPS (Treasury Inflation Protected Security) yield as a proxy for the real interest rate, since the TIPS yield is effectively the nominal yield minus the expected CPI. A chart of the 5-year TIPS yield is displayed below and discussed in the next section (in relation to gold).
The TIPS yield is not an accurate reflection of the real interest rate because it is based on the CPI and because the market’s expectations are sometimes wrong, but for practical speculation purposes it seems to be good enough.