UK equities have risen more than 10 per cent so far this year, raising the question: how much further can the rally go?
There are reasons for optimism. Slight pick-ups in M1 growth in China and the eurozone point to both economies recovering later this year; US-Chinese trade tensions seem to be waning; and a lack of wage inflation points to the US economy’s non-inflationary expansion continuing.
Equally, though, there are dangers. Inverted yield curves in the UK and US suggest there is a risk of recession; stock markets are pricing in cuts in US interest rates and so might be disappointed; a no-deal Brexit could plunge the UK into recession (although the market has so far been resilient to Brexit uncertainty); and tensions with Iran might escalate, pushing oil prices up.
These Jackanory approaches to forecasting won’t do, however. They invite us to commit a basic error – the narrative fallacy, a tendency to believe neat simple stories that gloss over messy reality.
Instead, we should ask: what is the market discounting, and what isn’t it?
Of course, the random walk hypothesis says that all foreseeable events are already in the price and so future returns will be random and unpredictable. But this is a hypothesis to be tested, not a foregone conclusion. If instead bad news is mostly discounted and good not, the market is more likely than not to rise if we get average luck. Conversely, if it is the good news that is already in the price, equities are in danger.
What we need, therefore, are indicators of investor sentiment that tell us whether good or bad news is in the price and which therefore can predict future returns. Although the random walk hypothesis says there should be no such predictors, history speaks otherwise. A handful of lead indicators, based on measures of sentiment, have predicted annual returns recently. They correctly forecast that the All-Share index would rise slightly in 2017 and fall in 2018. And at the end of last year I wrote that they are “now pointing to the index rising just over 10 per cent by next December”. Although a lot can change in the next five months, that forecast is on course so far.
These indicators point to the rally continuing. They are predicting a 15 per cent rise in the All-Share index by next summer.
Two factors in particular are bullish.
One is simply the dividend yield. At 4.1 per cent this is well above its post-1985 average of 3.6 per cent. This points to above-average returns.
The other is that non-Americans have been big sellers of US equities in the past 12 months, although their selling has diminished a little recently. This has in the past been a signal that they are bearish about equities generally and has therefore predicted rising prices as sentiment recovers. Big selling in 2015-16 led to global equities recovering in the following 12 months, and big buying in 2000 and 2007 led to prices falling.
These two bullish indicators are mitigated by a bearish one. The ratio of global share prices to the money stock in developed economies is high. This implies that shares are an above-normal proportion of investors’ portfolios and that cash is a below-normal proportion. This poses the risk that any rebalancing of portfolios might be away from equities, which would cause prices to fall.
Net, though, our indicators are bullish, pointing to a less than 10 per cent chance of prices being lower next July than this if past relationships continue to hold.
Of course, there are caveats here. One is that these indicators only work in predicting returns over longer horizons. Their power comes from the fact that investor sentiment tends to mean-revert: when investors are gloomy they eventually cheer up which pushes share up, and when they are cheerful they eventually become less so which pushes prices down. In the short-term, however, pessimistic investors can become even more pessimistic and so a cheap market can get even cheaper.
Secondly, it’s possible that past relationships will break down. One concern I have is that big selling of US stocks might not be a sign of pessimism but rather of global investors rebalancing their portfolios away from US equities towards other risky assets. If this is the case, such selling won’t predict future price rises.
Granted, there’s no sign of a breakdown yet: a year ago our lead indicators predicted a two per cent fall in the All-Share index which is within a gnat’s crotchet of the actual 3 per cent drop. But it could happen in future.
Thirdly, we must never base our portfolios upon central case forecasts alone. We must always consider the range of outcomes, which includes a chance of shares falling – albeit a smaller than average one. Personally, although my equity weighting is higher than usual for the time of year, I still have a big cash position too.
Ultimately, all forecasts are tests of hypotheses – whether you know it or not. The hypotheses I’m testing here include the theory that we don’t need stories to predict the market; that the random walk hypothesis is wrong; and that some indicators have a stable-ish relationship with future returns. Time will tell us how strong these hypotheses are.
