When investing on behalf of our clients or in our multi-asset funds, we use three main building blocks – equities, bonds (issued by governments and the corporate sector) and property. Across the portfolios we manage the allocation to bonds is currently rather low. For those who follow the markets and adopt a value investment ethos (as we do) the reason for this will be clear. The bond market is expensive.
The word ‘bubble’ is somewhat overused; it seems to be rolled out whenever an asset has increased in value. The word is being associated with the current market for government bonds and many corporate bonds*. It is possible for assets to rise in value or even be expensive without being in a bubble though. Are we really in a bond bubble, and if so, where do we go?
80 Years Around The Houses
When a bubble bursts, investors lose serious money. A look at history shows us that it is certainly possible to lose serious money in real terms (after taking inflation into account) in the bond market. Mega-statisticians Dimson, March and Staunton show that between the early-30s and late-70s investors in UK government bonds lost over 70% of their purchasing power**. This wasn’t in a straight line; World War II provided some respite for bondholders. Nonetheless, investors in fixed income in an inflationary world didn’t do too well overall. Then came the 80s, 90s and 00s and the market apologised. Those that dared to invest in the high-inflation environment of the early-80s were rewarded by handsome real returns triggered by falling inflation, falling nominal interest rates, and eventually falling real interest rates. The journey has been a round trip that took some 80 years. Yields are now as low as they’ve ever been.
Which Way Next?
Of course I cannot say which way the bond market is going to go over any arbitrary time period. But with the bull market price rises of the last thirty years and miniscule yields on offer there are many who think that the bond market isn’t just bubbling away, it’s boiling over.
Rather than guess the direction of prices, I prefer to look at what value we get from buying at current market prices given the future cash flows from an asset. Unfortunately in order to do this it’s necessary to have a stab at what the future might bring for the factors that govern those cash flows. To make things tricky, I’m interested in real, after inflation, returns. For bonds inflation is really important, as those mid-20th Century investors found out.
I find it difficult to predict what I will be eating for lunch, let alone what the UK/European/US/Global economy will be doing several years hence. Given the forecasting record of most economists, it appears I’m not the only one. So for the analysis a pretty broad brush is required, and an acceptance that no one scenario is a dead cert. So let’s look at a couple of scenarios, using gilts as the subject matter^.
Scenario 1: Inflationary Times – The Printing Presses Turn
Inflation has been low compared to history over the last decade, but still positive. Central banks are creating money, which all other things being equal would cause inflation. My calculations show that with moderate inflation averaging in the 4-6% range, investors in the ten-year gilt at current prices could lose 20-40% of their money in real terms over five years, without any change in real yields. Real yields (roughly the yield on a bond less inflation) are very low at the moment in the UK when compared with history. If they moved upwards, losses would be worse.
Using a longer-dated issue, the 2055 gilt, shows that over 20 years the same average levels of inflation would lead to real-terms losses of 50-70%, similar to those seen in the mid-20th Century.
Scenario 2: Deflationary Times - It’s the Economy, Stupid
Ah, but we’re not going to get inflation are we? Central bankers don’t seem to think so, otherwise interest rates wouldn’t be so incredibly low and they wouldn’t be printing money to push bond yields lower. Fund managers in certain sectors also don’t think so^^. It is true that in the near term there are many deflationary pressures caused by the fallout from the global financial crisis. Chief is the lack of demand caused by families and companies paying down their debts (‘deleveraging’).
If inflation averaged 0-2% over the next five years, those ten-year gilt holders stand to gain 10-20% in real terms assuming no change in real yields.
Asymmetric Bet – Part 1
The two inflation bands I’ve described (0-2% and 4-6%) are more-or-less equidistant from the current 3.2% rate of RPI inflation. But the potential losses during inflation outweigh the potential gains during deflation for the ten-year bond holders.
That is because I have assumed that yields can’t go below zero. This is known as the ‘zero bound’, and if you think about it, it makes sense. Why would you lend your money to someone, only to receive less of it back in the future? You may as well stick it under your mattress, or bury it in the garden. Admittedly some institutions would have to have very big beds or gardens to make that work, but the point is a sound one. In Japan, where there has been persistent deflation for over a decade, ten year bond yields have not fallen below zero.
Given my acceptance of a lack of forecasting ability, I have no better idea than to assume that both inflation scenarios are equally likely. If you think that, the asymmetry of returns means that conventional bonds are not a good bet.
Asymmetric Bet – Part 2
I mentioned above that real yields might move, and that the real yield on a bond is roughly its yield less inflation. Currently the real yield on ten year gilts is around -1%. If that didn’t change, and the rate of inflation didn’t change, then that is the real return you would achieve buying at today’s prices.
We know inflation can change, and real yields can change too. In the 1980s they were positive, as investors demanded a return above inflation for their investment. Perhaps this is because they thought inflation would continue indefinitely and possibly rise. They are now negative, which has enhanced returns for bond investors. Maybe this is because buyers think low inflation is coming and will stick around.
On average, over the last century bonds have yielded a return 3 percentage points above inflation per year#. I am a believer in the power of ‘mean reversion’ in financial markets. This means that over reasonably long periods of time, the future will look much like the past for returns on major asset classes. It is important not to be too precise in this assertion as the world is full of noisy data. With that in mind, keeping in broad terms, we can see that in the past investors demanded a return above inflation for their bond investments. They are not getting that now, and won’t in the future unless we do get persistently low inflation. My guess is that investors will want some real return in the future.
In other words, I think that the balance of probability is towards real yields rising. Rising yields means losses for bond investors.
Is there an alternative?
Bonds are a traditional way for investors who don’t want to take much risk to get some return (preferably above inflation). We can also use them as a parking space for cash that we might want to use elsewhere in the future. Right now, the downside risks from inflation and rising real yields mean that conventional bonds don’t cut it for either of these uses.
Governments and companies offer another type of bond that have their face value and interest payments linked to the rate of inflation (index-linked, or IL for those averse to typing). IL bonds are much like ‘conventional’ or non-IL bonds but remove the variation of returns with inflation. So over five years, whether we are in inflation scenario 1 or 2 above you will get the same real return if real yields don’t move, a loss of about -5% in total. That might not sound attractive, but compared to losses of -20 to -40% on conventional bonds might start to sound like a better option. On the flip-side, we can’t ignore scenario 2. Index-linked bonds will underperform conventional bonds if low inflation plays out.
Again assuming that both inflation scenarios are equally likely, IL bonds are slightly more attractive according to their probabilistically expected return. An additional benefit that can’t be calculated is the peace of mind from the removal of one cause for uncertainty, inflation##.
There is another feature of IL bonds that makes them interesting, and that is the insurance element. Over the long term, paper money (or fiat) systems of currency have shown a consistent tendency towards inflation. Central banks are deliberately targeting inflation, a steady debasement of the currencies they control. While I may not have a view on where inflation is going over five years, over a few decades I think there is a strong possibility it will be positive, and at times potentially quite high.
The longer-dated conventional bonds mentioned above could lose 50-70% of their real value in moderately inflationary environments. Equivalent IL bonds would limit this loss, and have a high chance of a real gain if held for long enough. So I think that IL bonds offer insurance against an inflationary event that I think is quite likely.
In conclusion - What's hot and what's cool
I agree with the view that bonds are bubbling, and potentially boiling over. Where long-term savers want to preserve the real value of their investments, conventional bonds are not the place to be. Even looking at a shorter time frame of five years the odds of making positive real returns are not great.
Where savers want to protect the real value of their investments index-linked bonds represent a much better alternative. The inflation protection comes at the cost of potentially underperforming in a persistently low inflation environment. If you must own bonds, I think that is a risk worth taking.
5th December 2012
*See for example the FT, Monday 21st November 2012
**Dimson, Marsh & Staunton, Credit Suisse Global Investment Yearbook 2011
^Bond market data from Bloomberg and Factset. Calculations are my own.
^^See for example “The bond bubble that never was”, Fund Strategy, 26th November 201
#Barclays Equity Gilt Study 2012
##Humans like a sure thing – I can heartily recommend Daniel Kahneman’s “Thinking, fast and slow” if you like psychology and/or behavioural finance.