This underperformance is more remarkable than it seems, because one thing should have given a massive boost to growth stocks - the fall in real interest rates.
Simple maths tells us that these matter enormously for growth stocks. To see why, consider a stock paying £5 of dividends which are not expected to grow in real terms. And let's assume that real bond yields are 4 per cent and the equity risk premium is three percentage points. The Gordon growth model tells us this share's price should be £71.40, giving a dividend yield of 7 per cent. It's a value stock. On the same assumptions about bond yields and risk premia, a stock paying £2 of dividends which are expected to grow by 3 per cent a year should be priced at £50. It's a growth stock.
Now, what happens if real bond yields fall to 1 per cent, all else equal. Our value stock should rise to £125, a gain of 75 per cent. But our growth stock should rise to £200. That's a quadrupling.
The logic here is simple. A growth stock - by definition - offers more cash flows in the distant future than value stocks. And a fall in interest rates means future cash flows are discounted less heavily, which should boost growth stocks relative to value.
This is, of course, no mere abstraction. Since the mid-1990s, real long-term interest rates have indeed collapsed. Given this, growth stocks should have massively outperformed value. But they haven't. Why not?
A big reason is that investors are prone to some cognitive biases which cause them to overestimate growth with the result that 'growth' stocks are overpriced and so subsequently underperform.
One of these biases is overconfidence: people exaggerate their ability to foresee future growth, and underestimate the extent to which it is in fact random.
Related to this is the planning fallacy. We tend to underestimate the extent to which things can go wrong - and for a small growth stock, plenty can. A glitch in its new technology might delay the product coming to market; trials might prove unsuccessful; as the company expands, managers might lose control of costs; backers might lose patience and not want to refinance the company; rival companies might produce similar or better products; and so on. The probability of any one of these many possible disasters might be small, but they multiply to produce a high chance of failure.
Yet another problem is herding. Because the future is uncertain, it's easy to take our beliefs about it from what others are saying and doing. This carries the danger that we'll buy near the peak of a bubble. It's no accident that the IT sector - a growth sector - has been so bubble-prone. Luckily, this danger can be reduced by following a simple rule: to sell when the price falls below its 10-month average and buy when it is above it.
A final problem is simply that some investors have lottery-type preferences: they want the 10-bagger rather than the get-rich-slow defensive stock. This too causes 'growth' stocks to be overpriced.
You might wonder why smarter investors don't exploit these biases by simply going short of overpriced growth stocks. Simple. They can't. Growth stocks can be volatile and (especially on Aim) illiquid. This makes it difficult and dangerous to go short of them.
Worse still, some investors might actually exacerbate the overpricing of growth stocks. Many of them cannot or will not borrow. This means that if they are bullish they don't do what the textbooks recommend, and take a leveraged position in the market generally. Instead, they buy high-beta stocks in the hope that they will rise even more than the market. Such stocks are often (but not always) growth stocks - so again, these become overpriced. As economists at AQR Capital Management have shown, it pays on average to bet against beta.
None of this is to say that growth stocks are always bad. They can beat high yielders in economic downturns simply because the latter tend to be more cyclical. Growth stocks beat value in 2008-09 as high-yielding builders got clobbered. And they beat value last year because China's slowdown devastated high-yielding miners.
However, it is by no means certain that growth stocks will get the benefit either of a cyclical downturn or of even lower real interest rates in 2016. If they don't, growth investors will find themselves betting against history - yet again.
