16 Jan 2018

Stock picking and the art of motorcycle maintenance

 “Why else would people pay fortunes for some things and throw others in the trash pile? Obviously, some things are better than others ... but what's the betterness? ... So round and round you go, spinning mental wheels and nowhere finding anyplace to get traction. What the hell is Quality? What is it?”

“The past exists only in our memories, the future only in our plans. The present is our only reality."

― Robert M. Pirsig, Zen and the Art of Motorcycle Maintenance

The futility of trying to consistently predict the future is well documented. Our emotional and behavioural biases coupled with the complexity of the world makes it virtually impossible.  In a highly respected essay, “Seven Sins of Fund Management” (Google it and you’ll find it), written by James Montier of GMO when he was at Dresdner Kleinwort, forecasting was sin number one.  James noted, “An enormous amount of evidence suggests that investors are generally hopeless at forecasting. So, using forecasts as an integral part of the investment process is like tying one hand behind your back before you start.” 

The problem with forecasting is it is essentially guesswork plus maths.  While the maths bit is relatively robust, there are all sorts of problems around the guesswork.  Our guesses of the future are based on a massive but incomplete information set and are generally biased by recent events, and can be subliminally anchored to completely irrelevant facts.

Even if you get past these issues, to think you can out-forecast the market for all the stocks in your portfolio for every result must surely make you guilty of sin number two, “The Illusion of Knowledge”.

Now, a recent study in the CFA Financial Analysts Journal, has found that even if you manage this, it doesn’t affect your return much. Feng Gu and Baruch Lev found that the gains from predicting corporate earnings, or consensus hits and misses - an activity at the core of most investment methodologies - have been shrinking fast over the past 30 years.  And that any residual gain that is left can be replicated by a “dumb” momentum investment strategy.

And yet, every six months, the industry moves into a frenzy of pre- during- and post- earnings analysis during the earnings results seasons of listed companies.  Models are updated, earnings hits and missed summarised, and management presentations held.

While you obviously need to review the performance of companies that you are invested in, trying to forecast the next six month’s earnings to the decimal point seems to us to be a bit of a waste of time.  More importantly, it leads to a focus on the short-term and a potential risk of not seeing the wood for the trees.  If you are focused on earnings and are worried the company will miss consensus earnings you might sell it, even if you see the long-term value.

The key to finding successful investments, in our opinion, is not to try to forecast what every company might earn in the future.  A better approach is to find a company that is so cheap on past earrings or assets that there is sufficient margin of safety for future hiccups, or a company of such “Quality” that is has good potential for success in the future.

So, what the hell is Quality? What is it?

While there are plenty of characteristics that might indicate a company of Quality, by far the simplest and easiest one is dominant market share.  Due to the competitive nature of business it is very, very difficult to achieve, let alone retain, a market share above twenty percent. To do that you need some sort of an enduring competitive advantage.  Any company with a market share materially above this is worth having a look at to see if it has one.  A competitive advantage can come in many forms.  It can be a brand (Coca Cola), strategy asset (Auckland Airport), network effect (Trademe), IP protection (A2), or scale (one of our two supermarket chains).

The problem with investing in these companies is that they are generally easy to see and the best monopolies are generally regulated to ensure they don’t abuse their commanding market position.  You are generally likely to do well investing in them in the long run, but you seldom get them at bargain prices. 

The next approach is to look for companies with the capability to grow into a market leader.  These are harder to find.  In his book “Common Stocks, Uncommon Profits” Philip Fisher provides a blueprint of how they might be found.  A meaningful profit margin, a culture of innovation and improvement, and the potential to grow revenues substantially are some of the characteristics that he notes as important.  A lot comes down to company culture, that fuzzy quality that is difficult to quantify.  In our opinion, a detailed financial forecast of these businesses will generally understate their value. This is because they are likely to do better for longer, mainly due to all of things they will do in the future that a financial analyst cannot predict.

A good example is a recent investment we made in Kogan.  Kogan is a portfolio of online retail and services businesses, focused on price leadership through digital efficiency. It listed on the Australian market in July 2016 with lacklustre interest.  It was undoubtedly affected by investor concerns regarding the imminent arrival of Amazon, making predicting earnings in the near term difficult. We focused on the qualities and strengths of the business and came to the opinion that, while they probably would be affected in the short run, Kogan business model would allow it to grow alongside Amazon.  More importantly we saw a smart and aligned executive team that would innovate and evolve the business.   This is important.  For instance, no one would have had Pet Insurance in their Kogan models two months ago.  This option value from innovative ideas is generally missed in the financial models of quality companies.

In our opinion, focusing on the characteristics of a company that makes it long-term Quality is not only easier but more financially rewarding than trying to predict the next earnings result.