There are two concerns with inflation-linked bonds, relating to how they are used as a hedge. The first concern is that they do not diversify equity market risk. The second concern is that they are an inefficient inflation hedge. These factors then lead to questions as to how they fit in with portfolio allocation strategy.
Although portfolio constructions vary, most investors have a heavy weighting in equities and other risk assets. If you need to meet retirement cash flows (either for yourself or for a pension fund), you want assets that provide a good return, to lower required contributions to the portfolio.
Furthermore, even the fixed income portions of portfolios tend to be filled with risky instruments. A typical way for fixed income investors to beat their benchmark is to be overweight credit risk.
The net result is that the bulk of the portfolio consists of risky assets. Although every pension consultant scours databases to find uncorrelated assets or investment strategies, all risk asset correlations go to one during a crisis. (If we strip out the jargon, they all drop like rocks.)
This is not surprising to anyone who has read the works of Hyman Minsky. When things go well, investors apply increasing amounts of leverage to their portfolios. When there is a reversal of sentiment, they need to start liquidating assets. Since everyone goes to the same conferences (and drinks at the same bars), they all end up owning the same assets. The selling frenzy eventually infects all the assets in their portfolios.
The role of conventional government bonds is to stabilise private sector portfolios. Their prices go up when everything else goes crashing down. Any investor that is overweight these bonds is in a great position to go bargain hunting. This buying support is what breaks the downward spiral in private sector asset prices.
Inflation-linked bonds’ weakness is that they too are in the “risk asset” class in the case of a deflationary crisis. This is exactly what happened in the Financial Crisis; investors in inflation-linked bonds were one of the groups that were carried out in stretchers after Lehman Brothers went belly up. This diversification issue is of paramount importance in portfolio construction, as described below.
Inflation-linked bonds might be useful as a hedge during an inflationary crisis, such as a repeat of the 1970s. Although market commentators have been predicting such an event for a very long time, such crises have been quite rare in developed countries. Furthermore, something like an energy equity position seems like a more interesting asset to hedge against such an outcome.
Another concern for inflation-linked bonds as a portfolio instrument is their efficiency as a hedging instrument. They certainly can hedge particular inflation-linked cash liabilities, as they provide similar cash flows. The issue is that they only provide inflation protection for the dollar amount equal to the value of the portfolio. As soon as we start allocating our portfolio towards other assets, that part of our portfolio is no longer indexed to inflation.
One way to deal with this problem is to apply leverage to the portfolio. The brute force method is to borrow to buy inflation-linked bonds. For example, if we borrowed $1 to buy $2 worth of inflation-linked bonds, we would have $2 worth of indexation protection for every $1 invested. The difficulty with this strategy is the realistic possibility that your borrowing rate will rise in line with inflation. If the real short rate you are borrowing at rises, you will end up eating a lot of negative carry, and that would cancel out your extra inflation protection. Furthermore, you have a leveraged portfolio, and are vulnerable to squeezes if lenders step away from providing liquidity (as in the Financial Crisis).
Inflation swaps are an alternative way of providing leverage. We can think of a swap as two counter-parties extending each other loans, so each ends up with 100% leverage. The advantage of an inflation swap is that the other leg of the swap is a fixed rate, so you have locked in the financing rate. This prevents the funding leg from cancelling out a rise in inflation. However, there are downsides.
- You are exposed to a fixed income derivative, and you need to be able to understand and manage the position.
- You need to provide collateral, and you will need to provide more collateral when inflation expectations are falling. That is likely to be in a bear market for risk assets – when you want to be able to buy risk assets.
- The counter-party risk of inflation swaps is elevated. If there is a persistent breakout in inflation, the value of the inflation leg will keep increasing, but the inflation receiver does not get cash flows to reduce this valuation gap. It may be unclear whether anyone will be able to make the final payments.
- The inflation swap market is presumably illiquid (I am not familiar with the current situation), and there is a huge bias by clients to receive inflation. It is quite possible that the pricing will be unattractive.
The other strategy is to find assets that provide greater economic leverage to inflation. During the 1970s, commodity prices rose faster than inflation. Therefore, a relatively small investment in a similar asset could provide inflation protection for the entire portfolio.*
The experience of gold after the late 1970s bubble shows the limitations of that strategy. Gold indeed rose more than inflation in the 1970s. Unfortunately, gold was a dead asset for roughly the next two decades, slowly giving up previous gains. Meanwhile, the CPI index kept rising. By implication, the price of gold fell in real terms, as shown in the bottom panel of the figure. In other words, gold did not act as a hedge against the realised inflation of the 1980s and 1990s.
One could argue that it might have hedged against a revival of the 1970s level of inflation. However, this would not help many institutional investors with inflation-linked liabilities. It is often the case that the maximum inflation protection offered by the private sector is limited in any particular year – capped inflation. This is most formalised in the United Kingdom, where there is a well-defined Limited Price Indexation (LPI) methodology – and in fact, LPI swaps can be traded.** Therefore, a very high inflation over a short period is no more damaging than sustained high inflation, which is the worst-case scenario. In other words, a sporadic hedge is not helpful in this context.
Real estate seems to be a more promising inflation hedge. That seems to be the message of institutional investors, as real estate is a popular asset class. I have some misgivings about real estate, but it certainly has been the best inflation hedge for institutional investors in the post-1980 era. Arguably, real estate is the main competitor for inflation-linked bonds in institutional portfolios.
The remainder of Section 4.8, which discusses how inflation-linked bonds fit within portfolio structures, will be published as another book excerpt.
Link to the Books2Read book page for Breakeven Inflation Analysis (has links to online booksellers).
* This was a suggestion I saw presented at my first inflation-linked bond conference; cannot remember the source.
** The usual definition of annual indexation increases under the LPI is the lesser of annual RPI (CPI) inflation, or 5%; indexation also has a floor at 0%, i.e., no deflation. However, since 2005, 2.5% can be used. A definition of LPI can be found at an online law glossary: https://uk.practicallaw.thomsonreuters.com/8-206-3992
(c) Brian Romanchuk 2018