Pig ignorant: Why blindness to base rates can hurt investor health


Download PDFGo 

Liam Nunn, European smaller companies analyst at Old Mutual Global Investors  

As I opened my fridge and instinctively reached for the packet of bacon, I suddenly found myself hesitating.

Saturday is usually fry-up day: after forcing some loathsome 'super-fruit granola' down my neck every weekday morning, I like to reward myself with a no-holds-barred full English at the weekend. But with recent high profile headlines proclaiming bacon is 'as big a cancer risk as smoking' still fresh in my mind, I retreated from the fridge and grabbed a bowl of All-Bran.

Was my behaviour rational? A classical economist would presumably argue that my actions fully reflect the utility I derive from consuming bacon, now properly adjusted for the risks. But I find it hard to believe I was solving a complex utility-maximisation problem on the fly (particularly that early in the morning, when I can barely muster the mental power to put my underwear on the right way round). It seems far more likely that I was fixating on a particularly scary statistic I had heard the previous week: namely that consuming fifty grams of processed meat a day can increase the risk of colorectal cancer by around 18%[1].

Indeed, as I munched miserably on that bowl of soggy cardboard, it dawned on me that I hadn’t given a great deal of thought to what an 18% increase actually meant. It sounded instinctively scary; but when I stopped to think about it, I had no idea what the chances of developing colorectal cancer were in the first place. Without the base rate probability of diagnosis, knowing there was an 18% higher risk didn’t actually tell me much at all. However, a quick Google search soon pointed me towards an article that helped spell out the risk in a more tangible manner: ‘The average risk for Americans of getting colorectal cancer is 5%... An 18% increase in that risk means the risk rises from 5% to just under 6%. That’s the absolute risk. The 18% is called the relative risk…both numbers are important, and yet the public rarely hears both of them[2].’

I don’t wish to suggest that the risk revealed by the World Health Organisation isn’t worthy of our attention – or indeed that it’s irrational for somebody to reduce the risk of such an awful illness in any way they can. I do think, however, most people would agree that seeing the conclusions of the study presented in these terms is significantly less alarming than the ‘18% increase’ the mainstream media chose to run with. By placing the new information firmly in the context of the underlying base rate (in this case, 5%), we get a much clearer view of the risk we’re taking when we opt for the occasional fry-up. 

In hastily abandoning my beloved weekend breakfast, I was guilty of something behavioural economists call ‘base rate neglect’ – a tendency to ignore prior probabilities in favour of more recent, eye-catching or readily-available information[i]. By failing to consider the prior probability, I found myself unduly swayed by visceral headlines. This is a trap investors all too often fall into when it comes to building portfolios. Numerous historical studies have shown that, on average, buying stocks on low multiples of earnings (or asset values) tends to generate significant outperformance over longer time horizons, while paying relatively high multiples for businesses is generally a recipe for long-term disappointment. 

However, despite the compelling evidence that our chances of success are higher when we build a portfolio of relatively lowly valued stocks, it's all too easy to ignore this prior probability when evaluating potential investments. When we come across lowly valued businesses, it's almost certain there will be some worrying headlines surrounding them; perhaps a profit warning, a delayed product launch or a management team that's losing credibility with investors. Amid the doom and gloom of the recently available (and often highly visceral) headlines, it's all too easy to ignore the base rate probability of outperformance from investing in a basket of lowly valued companies. Similarly, when we look at businesses trading on relatively high valuation multiples, it's all too easy to exaggerate the importance of positive headlines while ignoring the prior evidence that a portfolio consisting largely of such well-loved 'glamour' stocks is unlikely to outperform over the long run. 

Of course, this isn't the same thing as saying every lowly rated stock is destined to outperform and there will always be highly rated stocks that grow to justify their higher earnings multiples. But while careful analysis of individual businesses can hopefully help us identify the material exceptions to the general rule, long-term success in investing is still likely to boil down to playing the averages. By keeping the prior probability of outperformance in mind and ensuring valuation-discipline remains a key part of our process, we can hopefully increase our chances of bringing home the bacon for our investors.

 

[1]The Guardian, http://www.theguardian.com/society/2015/oct/26/bacon-ham-sausages-processed-meats-cancer-risk-smoking-says-who

[2] ‘What’s the absolute risk of eating bacon?’ by Karin Klein, The Los Angeles Times, [http://www.latimes.com/opinion/opinion-la/la-ol-bacon-risk-20151030-story.html]