I ran into this article on inflation hedging with inflation-linked bonds from FT Alphaville. I just scanned it quickly, but I think that the the approach used makes the topic unnecessarily complicated. I addressed the issue in my book Breakeven Inflation Analysis, and the odds are that I also over-complicated the analysis on the basis that I wanted to keep my book longer than five pages. This article is an attempt to reiterate my views in as short a text as possible.
The way to think about this is to not dive into the weeds of details of inflation-linked bonds, and go back to basics. We need to analyse the following premise:
I want a guaranteed high return over a particular investment horizon if event X happens.
The starting point is where the event X covers all possible future states, making the statement “I want a guaranteed high return over a particular investment horizon.”
You certainly can get something resembling a guaranteed investment return over an investment horizon, but the problem is with the word “high.” That “guaranteed” return is known as the risk-free rate in financial jargon.1 (Yes, there are always potential disasters, like a nuclear war.) If the risk-free return is “high” for you, great — you do not have to sweat the details of the financial markets.
You can try getting enhancements over the risk-free rate, but you are exposing yourself to some kind of tail risk. The usual way to do this is to take credit risk. Financial academics are busy coming up with “innovations” that appear to give these enhancements, with the tail risks being disguised from the people using those strategies (such as selling volatility).
(If you believe that you have a magic scheme to generate guaranteed returns above the risk-free rate, you might as well stop reading since you should be so rich that you have better things to do with your time.)
The Impossible: Event X not Covered by the Instrument “Design”
The next effectively impossible thing is to have the event X being unrelated to the payoff of the instruments you are trading. A significant portion of market commentary is discussions of what instruments might do well under different circumstances. Although I think this is a useful approach, we have to accept that we are only going to get “correlations” and not guarantees.
Possible: You Buy the Risk at a 1:1 Ratio
If the event you are worried about is that the price of something goes up, you can often buy instruments that give you 1:1 exposure to that something. Examples include the following.
- You can buy gold to get a 1:1 exposure to the price of gold.
- You can buy index funds to get 1:1 exposure to the stock market.
- You can buy inflation-linked bonds to get 1:1 exposure to the inflation index over the lifetime of the bond.
Since our interest is in linkers, I would note that there are some caveats: tax treatment, and coupon reinvestment. That said, the major investors in linkers do so in tax-sheltered portfolios, and coupon rates on linkers are currently pitiful.
But the key point to note in the above description: over the lifetime of the (inflation-linked) bond. People who analyse over shorter horizons expecting an inflation guarantee are unnecessarily killing innocent electrons.
The problem with this approach is the phrase “high return.” A 1:1 ratio means that your instruments just keep up with the price of X. You would have to invest 100% of your portfolio in linkers to get 100% inflation protection of the portfolio value. Given the expected returns on other assets versus the quoted (indexed/”real”) yield of linkers, that is an ouchie.
Possible: Buy X With Leverage
One could try to get the “high returns” by using the secret sauce of Scientific Finance: leverage. Although leverage does offer the prospect of high returns, you are exposed to the cost of financing the position as well as forced liquidation risk.
In the case of linkers, we saw how things could go horribly wrong with leveraged positions in 2008-2009.
Possible: Side Bet on X
The final way to hedge against a risk X is to undertake a bet with another market participant on the event. However, as soon as you do this, you are not getting an absolute protection, rather relative to market pricing. If there is a function CPI swap market, if it is pricing in 5% inflation next year, there is no guaranteed way to make money if the inflation rate does end up at 5%. (E.g., you could sell inflation volatility, but then you are stuffed if inflation ends up far away from 5%.)
Since there is no particular reason to care about the inflation rate only over the next three months, the only sensible side bet is a longer-term inflation swap — which is economically equivalent to a leveraged breakeven position, and thus has all the mark-to-market/liquidation risk concerns.
Linkers offer 1:1 inflation guarantees (modulo coupon reinvestment, tax) over the lifetime of the bond. Expecting an “inflation hedge” on any other horizon is a case of not understanding the terms sheet of the investment.
Those of a pedantic nature will note that what investments can generate the risk-free return are somewhat vague. People loosely talk about the Treasury bill rate as the risk-free rate, but bills can get expensive versus other instruments with no private credit risk. Since we are talking about a couple dozen basis points at most, this is a nothingburger for anyone not employed in the money markets.