What should we think about government bonds?
In a multi- asset portfolio, bonds can offer a variety of valuable characteristics: they can provide an ‘insurance policy’ (albeit an uncertain one) against losses in other areas, they provide a steady income stream; and they can provide potential capital gain to the extent that yields have room to fall. This is all well understood by bond investors. They can also lose lots of money. As with other assets, they are especially prone to doing so subsequent to a long period of price increases, or when investors become complacent about the likelihood of policy change. .
Today it is clear that bonds have had an exceptional run for over ten years, and that yields are extremely low. However it is lazy to assume that this means a poor return from here. How should we really think about these yields as investors?
Why are yields so low?
The major justifications for the where yields stand today fall into two broad camps. The first, that there has been a structural shift, suggests that yields are due to persist at low levels and should even fall, particularly at the long end. The second view is that today’s low yields are temporary, that we will see a “normalisation” in growth, inflation, and government policy, prompting bonds to sell off significantly from here.
Let’s look at each argument in turn:
The first argument – structural shift
The first argument is based on the view that major economies are “going Japanese,” with low growth and low inflation for many years; meaning that equities offer a poor and uncertain return.
This would mean that not only is it right to hold bonds, but that yields at the long end of the curve are still too high(!) for an environment where deflation is a bigger risk than inflation and policy rates are anchored to zero.
If this were true, then it would then be reasonable to expect long yields to revisit the lows of 2.45% (from 3.16% at the long end in the UK) and even lower (a rally to 2.45% from current yields would equate to a capital gain of 18%).
The second argument – policy, policy, policy
The alternative is that we are seeing a “normal,” albeit protracted, recovery in line with traditional economic cycles. In this world a pick up in growth would ultimately remove the need for accommodative policy.
If we conclude that long term inflation and growth will approach their long term averages, then we should also conclude that something close to “normality” will eventually resume for government bonds. This means that significantly higher yields should be expected at some point (higher than 4.5% for the UK 30 year gilt which, from current yields would equate to a 22% capital loss).
It also suggests that factors supporting higher yields such as materially stronger growth or indications of a shift in policy are likely to elicit a significant sell-off in bonds, potentially in pretty short order. This is what we saw at the end of May in response to Bernanke’s testimony to the Joint Economic Committee.
In short, either the consensus is wrong about long term growth and inflation and you can expect bond yields to fall as these factors fail to normalise, or there must come a time when yields must rise.
What do the “experts” say?
To what extent is it true that expectations for growth or inflation are materially lower? It is entirely plausible that bond yields are simply reflecting a cyclical shift in policy rates (itself in response to weak cyclical growth), with the market’s belief about longer term growth and inflation pretty much unchanged. Support for this can be found in a couple of places.
First, economists’ growth forecasts (surveyed by Consensus Economics) have shifted a bit lower for the medium term, but by no more than a coupe of tenths; and further out, forecasts revert back to the long run trend pretty quickly.
This is also the case with consensus inflation forecasts. Second, inflation expectations as measured by the yield gap between TIPS and Treasuries shows there has been no shift in 10-‐year inflation views at all over the past decade (ranging between 2.0 and 2.7% -‐ so a fairly narrow bound)
What do the facts currently tell us?
The US economy has been expanding steadily for the past few years, but the pace of growth has been sufficient to bring down the unemployment rate only relatively slowly compared with past recoveries. The modest pace of expansion is visible across comparisons of key indicators such as industrial production, consumer spending, employment and overall GDP growth in the current recovery with periods subsequent to past recessions (see figure 1 below). Indeed, compared with past ‘recoveries’, the recent experience has been woefully below par.
A notable exception to this has been corporate profits which, after a somewhat subdued start, have established an impressively normal-‐looking recovery path (see figure 2 below) – therefore taking a larger share of the growth that has been available this time around. Strong corporate profits out of modest growth is clearly pretty supportive of equity valuations.
The modest pace of demand growth – coupled with the unusually large amount of spare labour capacity which exists thanks to the depth of the recession – strongly suggests that a prolonged period of expansion would be required before any material inflationary pressures were generated. In that sense, evidence of cyclical inflation pressures should remain distant for some time yet.
This won’t stop expectations of inflation returning and nor will it necessarily prevent ‘pre-‐emptive’ policy action being taken (or spoken of) in the event that growth picks up further in the near term; which presents the challenging prospect of the early phase of both of the above scenarios (chaotic sell-‐off or range-‐bound buying opportunity) being indistinguishable.
So where do we stand on bonds?
A basic tenet of our approach is to acknowledge that we cannot consistently (or even occasionally) claim to know more than the market about macro fundamentals. We can, however, observe price behaviour in the context of changes in facts with the aim of identifying – and hopefully explaining – discrepancies. This gives us a chance of putting the odds in our favour if the market chooses to offer us good assets at a discounted price.
The key question is, how do the odds of losing more than 20% in government bonds (as yields rise to 4.5% and beyond) compare with the odds of gaining 20% if the yield falls back below 2.45%?
The lack of evidence for a structural shift lower in either economic growth or inflation expectations favours a reversion to more ‘normal’ yields when cyclical indicators (growth and policy) turn decisively. Nobody can tell when this will be, so it could feasibly be next week; or it could be in two years’ time. An observation of the macro fundamentals supports an extremely benign view on inflation for the time being, but we can have no real comfort that this is a view that will be shared by a market, whose structural beliefs on inflation aren’t all that different to where they were before the recession, amid stronger growth signals.
We should therefore be very wary of taking material exposures to US Treasuries against a backdrop of improving growth – such as that which can be generally observed at the moment. To be sure, while macro data are describing that improvement as being a mediocre affair, bond yields should remain in their lower range. But with the greater probability for markedly higher rather than lower yields at some point, we’d conclude that our exposures should be kept to a minimum.