The grim discovery that’s spooked the City

By Deputy Editor Tim Bennett Sep 19, 2011

Tim Bennett

Share with
friends:

Comments (5) Print this article

As a private investor, you’re unlikely to buy or sell enough shares to move a company’s share price significantly, at least outside the smallest stocks. In that sense, you are a ‘price taker’, in City parlance. The ‘price makers’ are the big institutions that trade vast quantities of shares.

But how do they decide what price a stock should be? For years, most City institutions, from fund managers to banks, have put their faith in one model above all others. It’s called the Capital Asset Pricing Model (CAPM). Never intended as a model for valuing individual shares, that’s nonetheless how it is widely used. What’s terrifying the City now is that, due to the credit crunch, the model is bust – its two key assumptions have collapsed. Worse still, no one knows what to replace it with. In the short term at least, that could be bad for share prices.

The origins of CAPM

CAPM helps investors assess investment risk and thus the price they’re willing to pay for an asset. This is crucial – since there is a direct trade off between risk and return, you must have an idea of what level of risk you are taking when you buy a share, or how can you know whether the returns on offer are sufficient? The model was seen as a breakthrough in modern finance when it arrived on the scene courtesy of William Sharpe – later Nobel Laureate in economics. He spelt it out in his 1970 book Portfolio Theory And Capital Markets. Although intended as a way to determine and deal with portfolio risk, it was quickly adapted as a way to calculate the risk attached to a specific share. Underpinning the whole theory are a couple of key questions that every investor must be able to answer.

See also

The death of the risk-free rate

Investing in shares isn’t done in a vacuum. When you take the decision to spend £1,000 on Tesco shares, you are giving up the option to put that money into a bank account, say, or other shares. This is known as the opportunity cost. So a key question for investors is: what is the minimum return I expect from a share? After all, shares are risky. A company could go bust and leave equity investors with nothing. The answer, therefore, has to be higher than the return you could earn for taking no risk at all.

But what sort of investment offers no risk? Until very recently the answer was a triple-A rated government bond (IOU). The assumption is that, as governments in stable Western countries can collect tax revenues (or print their own highly rated currencies), they will never run out of money to repay the IOUs they issue. That makes these IOUs pretty much risk-free. Since the biggest economy in the world is America, and it boasts the world’s reserve currency, it follows that the safest IOU will be a dollar-denominated US Treasury.

Investing in shares is a medium-to-long-term game and hence the benchmark for decisions about shares should be the expected return (or yield) from a medium-term – say, ten-year – US Treasury. So for years the CAPM model has built in the risk-free return based on what an investor in these IOUs could expect to earn. It then adds a bit to reflect the fact that stocks are all riskier than government-backed IOUs. But what happens when you can no longer assume US Treasuries are risk free?

That’s exactly the scenario now faced by investors after the credit-rating agency S&P recently downgraded US debt. As James Saft of Reuters says, “the words ‘permanent’ and ‘risk free’ have meaning and everyone will now know they cannot be applied to Treasuries”. That will make separating good from bad investment risks much harder, so institutional investors will build in a higher margin of safety and demand higher returns from equities. That’s bad news for share prices.

The death of the equity-risk premium

But the death of the risk-free rate isn’t the only problem. Everyone knows that shares are riskier than bonds, especially those issue by a triple-A-rated government, right? There are many reasons: dividend income is determined by a firm’s directors and not fixed as with most bonds. Shares are more volatile, so the risk of short-term capital losses is higher. If a firm goes bust (governments are not supposed to) bondholders get their money back ahead of share investors. So, there should be an extra return available if you take the risk of choosing shares over bonds. Indeed, a typical pension fund will assume it can earn a premium of anywhere between 4% and 6% for investing in blue-chip shares compared to government bonds.

Ten year rolling returns - equities v bonds 

The puzzle is that, for the past 20 years, “there has been no net equity risk premium”, notes Stephen Jen, managing partner at SLJ Macro Partners. “This has turned financial theory on its head.” For example, in the last ten years, US ten-year Treasuries have outperformed global stocks according to the MSCI world index (see graph). But if you can get a higher return from taking supposedly no risk than you can get for taking a lot, why take a lot? If the equity-risk premium is indeed dead, so are the prospects for many shares at the higher end of the risk spectrum.

Of course, what all this perhaps reveals is that the theories that have been propping up modern finance are sorely lacking. But that in itself suggests that we could be in for a difficult few years as investors adapt to a brave new world.

This article was originally published in MoneyWeek magazine issue number 555 on 16 September 2011, and was available exclusively to magazine subscribers. To read all our subscriber-only articles right away, sign up for a three-week free trial now.

Comments (5)

Share with
friends:

Comments

  • 1. Paul H

    (19 October 2011, 11:21AM)  Complain about this comment

    This is not really a new issue - there have always been theoretical (and not so theoretical) discussions about what is meant by risk-free, for example when setting regulatory standards. The truth is that nothing is risk-free, it is only a relative benchmark.

    The other point about economic theory goes back to a principle that you can't measure something without changing it. If a theory becomes believed by a sufficient part of the market, or a regulatory measure is imposed, then it will inevitably affect behaviour and reduce the validity of the data on which the theory was built.

  • 2. Tim Bennett

    (19 October 2011, 12:08PM)  Complain about this comment

    Paul H - I agree to an extent but recent events have revealed the full extent to which the City Emperor has no clothes, and arguably never did! Pricing shares is tricky for sure but how many people realise what a monstrous guess it is (and always has been)?

  • 3. Rob

    (19 October 2011, 01:55PM)  Complain about this comment

    This is indeed a significant issue. But what you fail to mention is that government bonds markets are now highly manipulated by central banks and governments. e.g. the Fed bought around 30% of new issuance in the past year through its QE programme, Bank of England is doing something similar. And now there is the "operation twist" in the US to lower long yields (the 10 year being key for CAPM models). Plus we have the eurozone propping up bonds in places like Spain and Italy with the EFSF etc.

    So not only are government bonds no longer "risk free" in terms of getting your money back (they still carry the risk of inflation taking off), they are also mispriced based on fundamentals due to massive official intervention.

  • 4. Rob

    (19 October 2011, 01:57PM)  Complain about this comment

    Part 2

    One answer is to take the inflation rate, add a bit for statistical fiddling by the government (if they say CPI is 4% then its probably 6% or 7%), add a couple of percent so you get a real yield (although still pre tax), then you get something closer to the real "risk free" rate. On this basis the US risk free rate is about 8%. Add 5% for equity market risk premium and the index should be priced with a cost of equity of about 12%.

    Since profit growth likely to be flat or weak for many years to come this points to a fair value market P/E of say 8 to 11. ie below current S&P level which is over 13.

  • 5. Mallard & Claret

    (19 October 2011, 02:51PM)  Complain about this comment

    Tim, Good article, however I am surprised by now that MW haven't written anything throughout this crisis on another model, which is quite simply far more dangerous to the financial system - Value at Risk. We missed the 'elephant in the room' again with the raft of new regulations making their way on how to make the financial system stronger and yet VaR doen't even get a mention. How on earth can risk be measured without capturing all the variables? Relying on neat models not fit for purpose is a recipe for further disaster and not the way and the regulators have missed their opportunity - again. I'm reminded of the saying "Those who don't know history are destined to repeat it"!

Leave a comment

This will be the name displayed with your comment.

This helps us verify comments are genuine. It will not be displayed anywhere on the site and is stored confidentially.

Please keep your comment within 1,000 characters and relevant to the main topic. We encourage healthy debate, but we don't allow insults or bad language. Anything off topic or unpleasant, we'll remove. Enjoy the conversation! Thank you.

captcha To prevent spam-related comments please enter the characters shown in the 'Captcha' box to the left.

By leaving a comment you accept our terms and conditions.


>