There’s a widespread consensus that equities will rise next year. At least one fact, however, challenges this.
This is the ratio of the broad money stock in developed economies to MSCI’s world equity index. My chart shows how this has predicted annual changes in the All-Share index since the 1990s*. High ratios of money to equities in 2003, 2009 and 2012, for example, led to big rises in prices, while low ratios in 2000, 2007, 2015 and last year led to shares falling. The correlation between the ratio and subsequent annual changes in the All-Share has been 0.55 since 1996, which is vastly more than would be the case if equity returns followed a random walk.
With the ratio now close to a record low – thanks to the recovery in global equities since March – it is pointing to the market falling next year.
One reason for this relationship is that the ratio measures investors’ portfolios. A low ratio of money to prices is a sign that investors have little cash but lots of equities. This means there’s a danger that they will try to rebalance their portfolios away from equities and towards cash. Such selling pressure would force down global equities, dragging the All-Share index down with it.
Another explanation is that this has nothing to do with money at all. The ratio of share prices to any upward-trending variable can predict returns. This is because stock markets sometimes overreact, rising too much or falling too far. A ratio of prices to money (or to anything else) tells us when this is happening and so tells us when equities are cheap or dear. And right now, this ratio is telling us that the market has overreacted to the good news about the roll-out of a vaccine.
However you explain the correlation, the inference is the same. It’s warning us that the All-Share index will fall next year. If the post-1996 relationship continues to hold, the index will fall 8 per cent, and there’s only around a 25 per cent chance of it rising.
Should we trust this message?
One reason to think not is that cash pays pretty much no interest, which means investors have less incentive to shift into it. A low money-price ratio might therefore be sustainable, and not a portent of a shift from equities to cash.
To test this hypothesis, I asked: how well has the money-price ratio predicted equity returns if we control for interest rates (as proxied by the US fed funds rate)?
And the answer is: it still does a good job. In fact, a lower fed funds rate alongside the money-price ratio actually predicts lower equity returns. This means that unless investors react differently to zero interest rates than they do to near-zero ones, we cannot rely upon low rates sustaining demand for equities over the next 12 months.
Another possibility is that the UK market could hold up well even if investors switch out of global equities, simply because UK shares are now so cheap. Personally, I doubt this. It’s rare for UK shares to rise in a 12-month period in which global equities fall much: this is only really possible if US big tech falls significantly while other stocks hold up.
There are, however, better reasons to hope that equities can rise. They start from the simple fact that this lead indicator is fallible – just like all other lead indicators and all pundits. In the mid-2000s and in 2017, shares did well despite the money-price ratio being low. The same thing could happen again.
One reason it might is that there has for years been a strong correlation between annual equity returns and annual growth in industrial production. This tells us that shares do well as the economy expands even if investors foresee the upturn. This could be because while investors anticipate growth in earnings and GDP, they don’t anticipate that such growth will increase their appetite for risk: instead, they project their current tastes into the future.
Also, this ratio is of course not the only lead indicator. Two others are actually sending bullish signals: the fact that the yield curve is upward-sloping, and the fact that retail sales are high relative to the All-Share index. Unfortunately, though, these indicators have been more successful as predictors of three-year returns than of annual ones. But we shouldn’t discount them entirely as a guide for the next 12 months.
So, yes, the signal from the money-price ratio could be wrong. But will it be very wrong? And can we be confident it will be wrong? (These are of course two different questions.)
I’m not at all sure. The ratio is warning us not to be swept along by the bullish consensus – because this consensus might already be in the price. This isn’t a reason to dump equities altogether. But it is a reason to hold a balanced portfolio in which we have some diversifiers against equity risk. We might just need them.
*I’ve adjusted the ratio to remove a steady uptrend in it.
