Join our community of smart investors
OPINION

The need for ground truth

The need for ground truth
June 23, 2017
The need for ground truth

Many pundits failed to foresee the election result because they cleaved too strongly to conventional ideas - such as that elections are won from the centre ground - and failed to appreciate sufficiently the ground truth that young people would turn out to vote in droves and that workers were cheesed off by years of stagnant real wages.

Investors face the same problem as political pundits. To take just two recent examples, the ground truth was that British Airways' IT systems were inadequate and that customers were shunning DFS. A grasp of these ground truths would have saved us money.

We can't rely on company accounts to alert us to such truths. What looks like good cost control might in fact be skimping on essential maintenance spending, which has nasty consequences. IAG (BA's parent company) is the latest - and relatively mild - example of a long tradition of what looks like good management in fact being exposure to tail risk. For example, in the 1990s Equitable Life earned a high reputation by making big payments to pension policyholders, but at the expense of having sufficient reserves to meet higher costs. BP under John Browne was seen as having good control of costs, until an explosion at the Texas City refinery revealed dangerous underspending on maintenance. And banks made nice profits in the mid-2000s through what Nassim Nicholas Taleb called picking up pennies in front of a steamroller - which worked until the steamroller accelerated.

Back in 1945, Friedrich Hayek said that economic knowledge is never fully possessed by a single individual, but instead consists in dispersed and fragmentary titbits across thousands of people. This, he thought, was why centrally planned economies couldn't work. But companies are forms of centrally planned economy, too. Just as economic planners cannot know the whole ground truth, nor sometimes can managers - and certainly not outside investors.

Hayek thought the solution to this was to use free markets: these, he thought, aggregated the dispersed knowledge of individuals. Prices, he thought, revealed ground truth.

But this isn't always the case. Some knowledge isn't embedded into share prices because insider trading is illegal, because many investors are unable or unwilling to short sell sufficiently, or simply because markets (for example in some Aim stocks) are illiquid. And prices can embody not hard knowledge but ignorance or irrationality. Brock Mendel and Andrei Schleifer show that credit derivatives were overpriced in the mid-2000s because traders bought them in the mistaken belief that they were fairly priced when in fact they were too expensive. The traders were, they say, "chasing noise".

Nor can we safely assume that the market selects in favour of companies that have control of ground truth. For one thing, markets can select not the wisest managers but the luckiest. Highly geared companies, for example, thrive when interest rates are low and the economy is growing but suffer horribly in recessions. And for another, markets select only for past success and in a changing environment this is no guarantee of future survival.

Investors, then, have a problem. We need to know ground truth to avoid profit warnings or warn us of impending financial crises, but we cannot rely on markets or management to reveal this truth in a timely manner. So what can we do?

One possibility is scuttlebutt investing: we get our shoes dirty by visiting companies and talking to workers, suppliers and customers about how they are really doing.

This isn't foolproof, though. What we learn might not be representative; anecdotes are prone to a selection bias. Instead, we might get only an illusion of knowledge that fosters overconfidence.

Instead, what we need is humility. Share prices, like election results, are the result of complex emergent processes. As Brian Arthur of the Santa Fe Institute says, such processes are inherently ever-changing and indeterminate - which means they are unpredictable. What I found so distasteful about the political pundits was not that they were wrong - everybody's wrong lots of times - but that they claimed a certainty that they couldn't (and didn't) have.

This has practical implications. I hold tracker funds not because I believe stock markets are efficient, but simply because I don't know the ground truth about particular companies and so don't trust myself to pick stocks. Equally, I'm not sure fund managers can reliably possess such truths, which is one reason why I prefer not to pay them high fees.

It also means there's a case for holding cash even at negative real interest rates. Cash limits our losses to the negative real rate. And it protects us from the worst that complex processes can sometimes produce, such as crises, illiquidity and correlated falls in asset prices.

Also, although we can't know what we're doing or foresee very much, we can at least avoid the most obvious mistakes such as trading too much, holding on to losing stocks in the hope they'll come good, or being overconfident.

By the same token, even if we can't know sufficient ground truth to pick particular stocks we can at least back some strategies. We know, for example, that momentum and defensive investing work.

But there's something else. We must ditch the pursuit of perfection. Participants in our portfolio clinic sometimes ask: "What am I missing?" The answer is often: "Something, but I don't know what". Where knowledge is unavoidably absent, we cannot optimise our portfolios. All we can do is what Nobel laureate Herbert Simon called "satisfice" - make do with a portfolio we are comfortable with. Mistakes and missed opportunities are inevitable, but we should ensure they aren't too costly, and if we have a balanced portfolio they shouldn't upset us.