The market is cheap (but specify your inputs)

Lots of people think the market is dangerously expensive. They take a look at the S&P 500 graph, which looks like this:


And they’ll point out that we’re now about 50% higher than we were in 2007, just before the last crash. Has there been a stupendous improvement in the real economy since then? No, they’ll argue, and so the index level must have been frothed up. You’ll probably hear something about printing money at this point.

But the graph in itself doesn’t tell us a great deal, since it’s raw data without any comparator. Much better to look at how the companies making up the index are doing, and how they are valued in aggregate relative to a proper measure, like earnings. At this point the Shiller PE graph often gets trotted out. It is basically a ‘normalised’ price-to-earnings graph, which shows the current multiple you’re paying for the average last-10-years earnings for the companies making up the index.


And, as you can see, we’re now at 27 – about where we were in 2007. But this is not, in itself, evidence that the market is ‘overvalued’ – at least in the way I define the word.

Consider a company which earns £10m every single year, with no volatility. Every year, it churns out £10m in cash flows, and it pays them back to its owners. Let’s pretend that everyone thinks there is no risk in the investment. How much is it worth? The answer, clearly, is that there’s not enough information to say. The asset I’ve just described is equivalent to a government bond or, for the everyday person like you and me, a savings account. If my savings account is paying me 10% interest, I would value the asset I just described at £100m (since at £100m, it’d pay me £10m yearly, a 10% yield – equivalent to my savings account). If the best savings account I can find is paying me 3% interest, I would value the asset at £333m (which would give me a 3% ‘yield’ on my investment).

Same asset. Two different environments. In the first environment, it is worth 10x cash flows. In the second environment, it is worth 33x cash flows.

This is not a ground-breaking analysis. In fact, it’s completely obvious when you think about it. The principle is that assets have to be priced relative to what else you can do with your money. Markets are predominantly driven by large financial institutions and not retail investors, and the asset alternatives the big money managers have are basically money-market instruments and long-term bonds. Hence, the price of these is absolutely integral to what an equity is ‘worth’. No man is an island – and by participating in any market you are at the whim of other investors’ preferences.

You don’t need me to tell you what the yield on those securities has done over the last 10 years. I think broad classes of UK equities are incredibly cheap in comparison. You might have seen evidence of this in your own behaviour or the behaviour of others – how many people you know have become bored of earning 0.5% in their bank account and have suddenly become interested in equities, looking for better returning assets? This is not a sign of a market bubble – it is a natural response to a massive deterioration in returns on assets people are more used to holding.

Now the riposte would be; “But Lewis, we’re not in a normal market environment – all you’re really doing is saying that because interest rates are so low, equities are cheap. If you were to start discounting equities using the interest rate now instead of a ‘normal’ interest rate, you’d clearly end up with an incredibly high valuation!” This is true. But I would say this; you shouldn’t conflate your opinion on what interest rates should be and how attractive you think the equity market is. Keep those opinions separate, and value things relatively.

Really, equity market valuations are up of three main ‘variables’:

– The stream of cash flows coming in
– The risk free rate (which serves as your ‘benchmark’)
– The equity risk premium (which is a difficult to measure number, but basically shows how much extra return investors demand for holding the higher risk asset class)

If you think interest rates at the moment are unsustainably low – that’s perfectly fine. But there are lots of ways to express that opinion by itself. Saying the equity market is expensive is expressing an opinion on all three of the things above. 

What concerns – or interests – me more than anything else isn’t fundamentally what I’ve just described. What I am surprised about is how few people are talking about these issues* – because they strike right to the heart of valuing companies. People cling onto P/E ratios are benchmarks of valuation because they are simple and because they’re used to using them. But assumptions about the correct discount rate to use when valuing a company are absolutely integral to deciding what a ‘fair’ P/E multiple is.

But things have changed, and to cling on to old ways of valuing companies isn’t conservative – it’s just lying to yourself about the assumptions you’re making. There is a warm, comforting safety in using the P/E multiple you’ve always used. Utilities are worth 14x. Risky small caps are worth 6x. But these numbers are just concoctions of variables baked up in a different environment, which may have approximated a fair valuation under a certain set of parameters. Some people will dwell on this and decide that, on balance, actually they’d really rather stick with that line of thinking. Perfectly reasonable – I just think more people should be having the discussion.

In short, what I am trying to do to make sure I understand the assumptions I am making is:

– Think about how interest rates affect what you think a ‘normal’ multiple for a business is. We are all used to valuing X type of company at a P/E of Y, and A type of company at a P/E of B. But a P/E is just a shortcut to a DCF. What does a multiple really mean? 

 Consider how interest rates will change the comparison between companies. Some of the ways it will do this are obvious – like pension schemes. Companies with big pension schemes will have big valuation changes depending on the level of interest rates. But there are also valuation quirks. Consider that blue-sky, growthy companies look much more attractive in a valuation model if you put in a low interest rate to discount it back (cash out now, high future cashflows discounted back very modestly)

– Specify what I really think about the macroeconomic environment. I am a bit worried that interest rates are low and equity risk premiums are low in certain sectors – think ‘bond like’ equities. This might be a neat way to hedge market risk. In fact, if I really think that the market is very cheap in the current interest rate environment, I could just go hard on the market and try to hedge off interest rate risk.

– … and, frankly, multiples are approximations at best in any case. They are a neat way of explaining an investment thesis, but they should not be what you hang your hat on. The understanding is important – since multiples drive markets – but they’re a shortcut, not an ulimate cause.

* Save, of course, Warren Buffett, who recently said that he thought the equity market was ‘extremely cheap’ if interest rates stay around here. He also said that, if it was practical, he would have shorted huge numbers of bonds to buy equities.

5 Replies to “The market is cheap (but specify your inputs)”

  1. cig

    Good point well made.

    If rates remain low(ish) long term, as seems likely to me, the penny will drop with the mainstream sooner or later. This should be a fun ride!

  2. UK Value Investor

    Hi Lewis, an interesting set of points. I don’t think there’s anything wrong in saying the S&P 500 is expensive based on CAPE, as long as you define expensive as being a certain degree above CAPE’s long-run average (or however else you might define it).

    So as you say, define your inputs.

    I think you’re also right that saying utilities are cheap at a PE of 14, or whatever, is nuts. It does indeed depend on what everything else is doing. If utilities are on a PE of 20 and everything else is on a PE of 50 then maybe utilities are cheap.

    I think in most cases relative valuation makes more sense than absolute valuation.


  3. John

    Perhaps they are cheap based on current interest rates because the market is factoring in future interest rate rises, and each expected cash flow is being valued based on the expected interest rate environment at the time that it is expected to be paid.

    • Lewis

      Yeah, that’s a good potential solution. Interest rates are tradeable assets, though, and there’s a few ways you can infer market expectations about interest rates in future periods.

      The easiest one, though, is probably that if you’re using a 10+ year bond as your ‘risk free rate’, you’re already encapsulating a lot of information about future rates in that number. Long-dated bonds capture a lot of the expectation of future interest rate movements in their price (otherwise you could just sell short-term bonds and buy-long term ones to arbitrage).

      For what it’s worth, UK 30 year bonds yield ~2.6%, so there’s clearly no expectation of much interest rate ‘normalisation’.

Leave a Reply

Your email address will not be published. Required fields are marked *