In April 2014 we published a special Market Tactics report looking at nine snapshots of UK equity valuation. Our prognosis for the FTSE 100 and FTSE All Share was for modest but respectable annualised real total returns on a seven year view. The strong positive correlation between UK and American equities was highlighted as the main risk to these projections, as valuations suggested worse long-term returns from the US market.
Fast forward to June 2015 and UK shares have delivered a solid performance, with the FTSE 100 total returns index up over seven per cent and the FTSE All Share TR index just shy of 10 per cent higher than when we published our 2014 report. Stateside, the S&P 500 has pushed on to fresh record highs and, with the dotcom peak breached in May and many credible valuation methods flashing red for the US, investors in the UK still need to be mindful of the risk that sentiment in the world’s biggest equity market could reverse. In addition, with the ever likelier prospect of Grexit sparking fresh turmoil in the Eurozone and a UK referendum on EU membership on the horizon, there are plenty of serious macro factors that could lead to volatility in capital markets. So, while acknowledging these significant risks, what do current valuations suggest about the outlook for UK shares?
In my update, I’ve chosen to focus on the measures for valuing the FTSE 100 (dividend yield, forward P/E ratio and price-to-peak earnings) and the FTSE All Share (dividend yield, CAPE ratio and P/B ratio) that have historically been most correlated with subsequent returns. In the original report, we used the real total returns that followed valuations in a similar range to forecast future performance. My approach to repeating this exercise is outlined in the box below but from the outset I should make clear that, in my opinion, the usefulness of the report metrics is in comparing how expensive equities are relative to the past - they do not provide an infallible guide to future prices and returns.
Forward PE ratio
Last year we noted that the FTSE 100 looked fair value according to the forward price/earnings ratio, which was around 13.3 at the time. That figure has now risen to 15.76 which is above the long-term average and (based on a median of annualised returns when the ratio has been between 15.61 and 17.47 in the past) points to a miserly 0.3 per cent real annual returns on a seven year view.
The forward P/E has been moderately predictive of future returns over the mid-term with a 36 per cent correlation between lower forward P/E and higher annualised returns (and vice-a-versa) over seven year timeframes.
Conducting analysis of shorter periods, the correlation is weaker but it is worth noting that from similar forward P/E ratios to now, the FTSE 100 has experienced negative real returns just one-sixth of the time on either a one or three year basis. On the more statistically significant seven year outlook, 50 per cent of annualised real returns have been negative. However, the worst was just -1.5 per cent, whereas the positive figures have in some cases been as high as six or seven per cent.
Looking at the different time horizons together, to me the forward P/E indicates that, while the current trend is towards the FTSE 100 becoming overvalued, that point has yet to be reached.
Dividend yield
Dividends make a vital contribution to overall equity returns and as yields directly imply the premium investors require to purchase shares at a given price, it is unsurprising that they have proven a very reliable indicator for large-cap shares in the past. Since 1986, overall dividend yield has been positively correlated with FTSE 100 returns on a one year (25 per cent of the time), three year (49 per cent) and seven year (73 per cent) basis.
At the beginning of June the dividend yield for the FTSE 100 was around 3.48 per cent. The subsequent performance when the DY figure has been in the range of 3.23 to 3.63 in the past implies three year annualised real returns in the region of 7.5 per cent. Over a longer seven year period, returns could smooth out to 3.6 per cent.
As with the less reliable forward P/E ratio, these figures are suggestive of further gains for the FTSE 100 in the near term.
Dividend yield has also been predictive of movements in the FTSE All Share. Although it does not stand out as the metric most correlated with total returns of the whole market, there has still been a positive relationship 34 per cent of the time on a three-year basis and 37 per cent of the time over seven years. With the current DY for the All Share at around 3.27 per cent, history suggests annualised returns of 3.8 per cent on a three year or 1.8 per cent on a seven year view.
Price-to-peak earnings
Another valuation that focuses on earnings is the price-to-peak method favoured by US manager John Hussman. The benefit of this method is that it takes into account that earnings fluctuate with economic cycles. In a recession earnings will be depressed, so even if share prices have fallen too the PE could still be relatively high, understating the value of potential future cash flows. In this scenario, if investors stay out of the market because of the high PE ratio, they could miss out on powerful share price rallies. Assessing the price of the market relative to peak earnings addresses this problem.
Currently the FTSE 100 is trading at just under 12 times its last peak in earnings which based on past results is suggestive of roughly 3.5 per cent annualised real returns over the coming seven years. This is a similar figure implied by the dividend yield method and in its own right price-to-peak has been predictive of 48 per cent of returns over this timeframe.
Interestingly, over a three-year period, the price-to-peak suggests much lower returns – 1.6 per cent - for the FTSE 100 than the dividend yield or the forward PE. As price-to-peak’s correlation with three-year returns is, at 25 per cent, half that of the dividend yield method, we should give it less weight as an indicator. It is interesting to note however, that given the methods predict near identical seven year returns, price-to-peak implies a more modest performance or possibly even a pullback sooner in the cycle.
CAPE
The cyclically adjusted price earnings ratio, which uses several years of inflation-adjusted EPS, has been shown to have more predictive power than a straightforward PE based on one year of data.
Using prices from the Datastream Total Market Index – a non-investable lookalike of the FTSE All Share – it is possible to compute CAPE ratios back to 1974. The long-term average is 15.05, so the current CAPE of 13.74 indicates that UK shares are not overpriced.
CAPE has been correlated with 30 per cent of three-year and 53 per cent of seven-year performance. Based on the median subsequent returns when CAPE has been between 12.57 and 15.05 in the past, the current level points to around 12.6 and 10.3 per cent annualised three and seven-year returns respectively.
These projections seem high but are reflective of the very substantial returns made by investors in equities who bought when CAPE was at similar levels to now in the middle of the 1980s (before CAPE spiked ahead of the October 1987 crash), the beginning of the 1990s (pre-dotcom) and, on a three-year view, at the bottom of the 2000-2003 bear market.
CAPE has been criticised as somewhat of a blunt instrument by institutional investors who are under pressure to outperform their benchmark. Professional money managers can look foolish if they sit on the side lines missing gains as they wait for the market to become cheaper. For example, the US market has been very expensive according to CAPE for well over a year but any manager who sold the US early would have had a hard time explaining to investors the forgoing of last year’s gains.
Of course for private investors there is no such benchmark risk but the danger of missing strong periods of upside means CAPE has its limitations as a market-timing tool.
Price-to-book
In the past, the measure of value most correlated with subsequent returns of the FTSE All Share has been the price to book ratio. The level of share price relative to the value of underlying assets has been predictive of three and seven-year annualised returns 41 and 80 per cent of the time respectively.
The All Share is currently trading on 1.96 times book value which is slightly above its long-term average. Based on subsequent returns when the P/B has been in the range between 1.82 and 2.14, the projections for the future do seem optimistic. Over three years the predicted annualised return is 16.1 per cent and over seven years it is 7.5 per cent. I am sceptical of these figures as they are heavily skewed by the tech mania of the 1990s. Since the turn of the millennium this level of annualised returns has only really happened in recovery from the lowest point of the 2003 and 2009 bear markets.
I wouldn’t argue with the price to book as a reliable indicator that UK shares are still decent value and could make gains over the medium term but given that we are quite some way from the last stock market lows, I would not expect to see the double-digit gains implied by the current PB over the next three years.
Dirty average
Giving each of the valuation metrics a weighting based on their past level of correlation with real annualised returns, the overall average projection for the FTSE 100 (based on DY, forward PE and price-to-peak), is for gains of 5.9 per cent a year over three years and 2.8 per cent annually over seven years.
For the FTSE All Share, the projection using the past figures for PB, CAPE and DY would be for roughly 11 per cent annually over three years or 7 per cent a year over seven years.
The mean annualised inflation-adjusted total returns for the FTSE 100 since 1986 is 6.8 per cent on a one year, 5.7 per cent on a three year and 5.3 per cent on a seven year basis. For the FTSE All Share the figures since 1965 are 8.5, 7 and 6.33 per cent for one, three and seven years respectively. UK equities have over the long-run been an excellent investment, so it is reassuring that, although not all the valuation metrics we have looked at suggest they are cheap, shares are not overly expensive by historic standards. So, while I would not necessarily expect to achieve the highest levels of returns forecast by my calculations above, I would still be comfortable buying some exposure to the UK equities market at current price levels.
Other considerations
Buying shares when they are reasonable value does not protect investors from the unpredictable events that spark major market sell-offs. The hope is that, having not bought too high, if markets take a tumble it will take less time for positions to recover in value and stay on track to achieve financial objectives over the longer term. In the past, when each of the valuation metrics looked at have been in the same range as current figures, there are variations in the amount of negative returns recorded over one, three and seven years. It is worth noting that the most predictive measures historically point to negative returns, including those of severe magnitude, being less probable than average right now. Although if history has taught us anything, it is that the improbable is certainly not impossible.
So, looked at in isolation, UK equities still don’t seem hugely overpriced by a number of measures. I certainly wouldn’t be out of the market but I would think carefully about the proportion of my total asset allocation in UK shares. Going a step further therefore, I would want to look at not only how valuations stack-up relative to UK shares in the past but also compared to other asset classes. From a strategy perspective, I also want to know what level of risk the proportion of UK equities holdings in a portfolio might pose to overall objectives.
The old adage “time in the market, not timing the market” rings true and valuations should not be used to decide when to dip in and out of investments completely. However, they can be a sign that tactical asset allocations need to be adjusted. These themes will be examined in the next two Market Tactics reports where I will look at portfolio risk-return optimisation and equity risk premiums.
