Experts advising Pension Funds have become like SAGE advising the UK government. Using historic computer based models based around some simple prior relationships, they make predictions on which they base their policy advice. The trustees, like ministers, are often non expert and thus defer to the advisors, who in turn advise/run an ultra cautious policy framework. For governments this ends up in a Zero Covid strategy. In investment terms this means holding too many bonds, since the advisor is only ever judged on the loss, not the potential gains foregone. The problem for both sets of experts is that the visibility of economic and social loss is increasing, as is the prospect of negative returns from bond funds.
The traditional Investment Committee meeting tends to start with a report by the Chief Economist (or some such) who will run through a laundry list of economic data points that everyone pretends are meaningful. After some general remarks about politics which lead to little or no conclusion, they move to the PMI (Purchasing Managers Index), which is supposed to tell us if we are in recession (it doesn’t) and has become something of a new favourite, taking over from the long time leader, the Non Farm Payrolls (NFPs), which is supposed to tell us what the Fed will do with interest rates – although that doesn’t work either.
As previously discussed, these high-frequency, monthly, data points tell us almost nothing about the real world but are useful in so far as they are the framework upon which short term traders hang their speculations. They are effectively akin to Sports betting, where positions are taken, bets are laid off, side bets and spread bets are added on and the whole thing is over in a short space of time. The role of the Economist is thus like that of the sports Tipster; to help generate the consensus around which the ‘book’ is created and to drum up noise and interest with a veneer of plausibility. And in the same way that Tipsters tend to both huddle together and to re-present what the ‘market’ is telling them as their own insights, so most of the short term data ‘forecasts’ tend to be ‘trend plus or minus one standard deviation’.
All good and harmless. Thus far. The problem becomes when it is applied to the real world (if we can call it that) of investing. Thus, prognostications on PMIs and NFPS translate into views on inflation and demand and theories as to whether to buy so called cyclical or defensive stocks, or indeed whether or not to buy bonds or equities. Views on the US$ drive views on whether investors should buy ‘Europe’ or ‘Asia’ and so on. But the reality is that the actual economic views themselves are rarely held to account and that in fact the economists tend to reflect back what they believe the ‘markets’ are apparently telling us, as well as what they think we want to hear.
This tends to result in the bond markets being declared wise and the equity markets stupid (economists tend to come from the Bond markets after all) and thus every time the yield curve (the difference between long term and short term interest rates) goes flat, there is supposed to be an imminent recession, meaning we should apparently buy long dated bonds and maybe gold and sell equities. Meanwhile, every time it steepens there is supposed to be inflation, which apparently means we should buy short dated bonds, index linked bonds and (if you really must) some cyclical equities. It also apparently sometimes means we should buy gold, as people who like gold will always find a reason to recommend buying it.
And this in fact is a key part of the problem, both for investment committees and governments; the principal/agent problem. The experts bring two unfortunate elements to the meeting, first their inherent biases and second their own risk/return framework. Thus the aforementioned gold bug will recommend buying gold regardless of the data, while the majority of economists will always want to buy bonds. If you already own a factor/style fund, like Value or Small Cap, having the fund manager in the room will rarely get an encouragement to sell their fund, either hold or preferably buy more!
Perhaps more important is that the agent – in this case the expert, but also the trustees to some extent – is judging their own personal career risk ahead of the returns available to the ultimate client. The famous ad that used to say that ‘no-one ever got fired for buying IBM’, was playing into exactly this fear. As such asset allocation rarely strays far away from the benchmark (the equivalent of buying IBM) and the Investment Committee minutes reveal the theatrics designed to ensure this. Investors not constrained in this fashion therefore need to avoid wasting time and effort replicating these theatrics.
For example, to take one of the key indicators used, the yield curve. The chart shows the yield curve of 10 year bond yield minus 2 year bond yield (10s minus 2s in the jargon) for the last 15 years as a dotted white line. The lower the number, the flatter the curve. The yellow, solid, line shows the relative performance of the S&P500 to the US Long bond, the latter as proxied by the ETF, TLT US. Thus, if our investment committee thinking is correct, then the lines should broadly move in a similar direction; as the yield curve steepens (white dotted line rises) then equities should outperform bonds (yellow solid line rises). And vice versa.
Chart 1. The yield curve is no longer helping us understand the economy
We can see an early example of how this didn’t ‘work’ in mid 2007, when the yield curve steepened sharply, but equities under-performed bonds. The opposite of what was supposed to happen. This tends to get lost in the noise around the 2008 Global Financial Crisis (GFC), but if we look carefully we can see that the yield curve was clearly telling us something a year earlier – just not about economics. Broadly, things returned to some form of normal for the following 5 years or so, but for the last 5 years – ie since 2016, things have tended to go in different directions from the theory once more. Except last year, when post the March intervention by the Fed, the curve steepened and bonds undeperformed.
After 5 years is the Yield curve relationship to Asset Allocation starting to reassert itself?
We can of course explain this in terms of the GFC itself, but more importantly in terms of quantitative easing and Unconventional Monetary Policy. However, in doing so, we need to also reset our analytical frameworks. Just as the NFPs no longer give us any insight into Fed policy, so the yield curve has stopped telling us about the economy. Indeed, the bond market appears to have lost its ability to give us any insight into the economic fundamentals and is now only telling us about itself. Useful, but very different from before.
There is also the issue that post 2009 China started to dominate a lot of global demand such that the US PMI started to lose its usefulness as a predictor of demand and, indeed, appears to have started to move in the opposite direction from where the ‘theory’ said the bond market should go since late 2015. And getting where the bond market is going is supposed to be the end goal, not simply getting the forecast of the indicator correct.
As a result, too many investment committees have become rather like the SAGE committee advising the UK government on Covid-19; economists who are specialists in Modelling rather than investors advise trustees (who are also rarely investors) on what to do, based on their models of the indicators for what will happen in markets. No one questions how accurate the models have been, nor whether they apply in current market circumstances. Perhaps most important, no one questions the validity of the policy recommendations being made – even if by some chance the forecasts were accurate and the indicators themselves still relevant.
Governments and Trustees are thus at similar risk from a form of expert ‘producer capture’; having delegated the risk decision to the modelers and hidden behind their credentials as experts they themselves become complicit in the experts’ personal risk aversion embedded in the policy recommendations. The opportunity cost ( a concept well known to economists, if rarely acknowledged by politicians) is what French writer Frederic Bastiat referred to as ‘that which is unseen’ in his parable of the broken window and for pension funds this is the forgone upside from a 10 year equity bull market. For politicians, that which is unseen, so far at least, has been the huge hidden cost of lockdown.
With US Bonds yielding around 1% and German bunds still guaranteeing a small loss on redemption the question is when that position starts to be ‘seen’. We think it is already starting to come visible, as capital losses from yields rising have little cushion from actual yields paid out. Certainly investors not already constrained by the committee process and its benchmarks are hardly rushing to buy bonds – except perhaps Chinese ones. For politicians, they should perhaps bear in mind that often the catalyst for a necessary change in investment policy is a change of fund manager and/or economic expert. Necessary, lest it become a change in trustee.
