According to Benjamin Graham “in the short-run, the market is a voting machine but in the long-run it is a weighing machine”. Put another way, the market is never wrong for long, but investors often are.
Currently, the S&P 500 Index is at an all-time high and according to most valuation metrics – such as historic or forecast price-to-earnings (PE) as well as the Shiller cyclically-adjusted PE ratios – the market is ‘very expensive’.
But does this mean the market will correct this year? We believe these metrics are never a sufficient enough argument to estimate expected market returns. At best, PE ratios might help us over cycles of five years or longer, which is not good enough to keep our reputations and track records in place if we get it wrong.
The market is constantly forecasting earnings per share (EPS) growth for us, but is anybody listening?
The key problem in using PE ratios is that they require somebody to assume what the ‘fair’ or realistic price-earnings level should be right now. In addition, PE ratios suffer from substantial problems and ‘dimensional biases’.
For example, stock prices change every day, whereas EPS is only reported twice or four times a year. Prices are far more volatile than EPS, so the contribution of price changes to the PE ratios by far dominates any contribution made by EPS changes. Mathematically it is accurate to state that a PE ratio is actually just the price with a tiny ‘earnings limp’.
Also, simple ratios like the PE can easily hide vital information.
If the PE of the market moves from 20 times to 22 times, this could either be due to prices rising or EPS falling and each means something very different.
What doesn’t matter … and what does
As a result, we have developed an indicator that can remove a lot of this subjectivity. We base our analysis on the old economist’s adage that using the ‘level’ of a variable like PE ratios actually never matters, whereas measuring ‘changes at the margin’ to these levels does.
We prefer using a type of ‘deconstructed’ PE ratio where the ‘rate of change’ in prices (in other words, how the market is reacting to new information at the margin) is compared with the ‘rate of change’ in EPS or earnings growth.
For example, if the market prices are rising by 15% per annum but earnings are only rising by 5% per annum, that is telling us something. Rationally, that is not sustainable.
Sometimes, in extreme cases, market prices may even be rising while earnings growth is falling. This demonstrates a clear market conflict, irrespective of what the current PE levels are.
Figure 1: S&P 500 EPS momentum vs price momentum (both year-on-year%)
We have therefore discovered that the rate of change of prices is a remarkably reliable predictor of future EPS growth – far more reliable than consensus EPS growth forecasts.
It is like the market itself is constantly and effectively estimating what future EPS growth is.
In the chart above, we see the rolling 12-month change in the S&P 500 excess return above US Treasury bills together with the 12-month change in the earnings of the S&P 500. Instead of combining price and EPS into a PE ratio, we realised that price momentum and earnings momentum don’t deviate too much from each other for too long.
Prices cannot rise sustainably without EPS growth eventually materialising.
For example, during the period 1997-2000, year-on-year price momentum was sustainably higher than EPS momentum. When the EPS growth did not materialise, the price momentum collapsed, resulting in the dot-com crash.
As Benjamin Graham stated, the ‘collective market machine’ is indeed a clever instrument that can more reliably ‘weigh’ what future EPS growth is going to be.
Figure 2: S&P 500 price momentum reliably leads EPS growth by six months (both yoy%)
This chart shows that if we lead the S&P 500 Index price momentum by six months we get a much better ‘fit’ than the same time series in Figure 1.
This indicates the market does to some extent effectively ‘forecast’ EPS growth.
So how do we use this information in the case of the US and the S&P 500 Index right now?
The interesting result of our analysis is that if EPS growth follows price momentum, as our research observes, then the market itself is currently expecting EPS growth to move from 4% year-on-year and to balloon to 18% over the coming months.
The market doesn’t care what you think
In conclusion, as we stand in January 2020, the market itself already ‘priced in’ during the 2019 market rally a rapid and extreme spike in EPS growth for 2020.
This can only resolve itself in two ways:
- Either the US equity market delivers abnormally high EPS growth in order to justify the run-up in prices, or
- Prices correct when S&P 500 corporates fail to deliver this very extreme EPS growth expectation.
We therefore conclude, without having to rely on current PE levels or using EPS forecasts from the experts, that investors can expect a pull-back in the US stock market in the first half of 2020 as the most likely outcome.
This isn’t our view or forecast – it is the market itself that has got itself into a knot. Price momentum has simply run ahead too fast from its own earnings momentum and generally this is not a good signal for positive equity returns over coming months.
Remember that the market is self-correcting and therefore is never wrong for long and it certainly doesn’t care what current PEs are, or what you or I think.
Roland Rousseau is responsible for equity structuring and portfolio strategy at RMB Global Markets.