Wednesday, July 2, 2014

Is There A Better Way Of Stock Picking?

How do you pick stocks? What method do you use to pick your stocks? Analysis reports? Broker recommendations? Forum tips? Insider tips? Personal experience? Fundamental metrics (P/E ratio, dividend yield)? Horoscope? Guessing? What makes your stock picking method better than everybody else?

I will address that question after I digress a bit, but don't mind me, it is a useful build-up to my final conclusion.

I have to say, Millennial Invest is one of the best blogs that I follow. His posts are relevant, informative and credible. I like his style of writing as well.

Strangely, a lot of his recent posts struck home with me, particularly this one, regarding "Over-diversification". Now, over-diversification already sounds like an oxymoron, doesn't it? People should want to be diversified... or do they?

I think the correct answer to this question is: What kind of diversification? Diversification of risk factors is my personal favourite, which is why I like to look at asset classes on the whole, since they all have different risks that affects them to different degrees.

But since we are talking about stock picking, we are exclusively within the asset class of equities. How do you diversify properly? The conventional wisdom by Bogle and Ferri is to do index investing. Buy the whole lot since you won't be able to figure out the winners or the losers. However, there has to be a better way right? And that is where I find his post so relevant.

Patrick took 4 factors - Value, Momentum, Shareholder Yield and Quality and split them to deciles and showed their performance. It looks a little like this:


Value is defined as paying less than the real worth of something. The problem is that the real worth is never a fixed point and calculating comes with lots of estimates. This brings about a lot of errors. The concept is easy to understand though. If you buy something less than it's fair value, it has the potential to appreciate back to it's fair value. The typical metrics are P/E and P/B, but there are others as well.

Momentum, a.k.a. momo, is not an easy thing to measure. The results are compelling though, but I don't really know of any good way of measuring momo, especially for stocks that pay dividends, since it breaks up their chart and skews technical analysis.

Shareholder Yield is an interesting one, and I must personally say that I love it as a factor. I have talked about it after I read Meb Faber's book on it. Shareholder's Yield = Net stock buyback + net debt repayment + dividends. Basically, it calculates the value that you receive as a shareholder, regardless of capital gains or dividend gains. This is a factor that I am very bent on using for my future analysis because it is very logical and intuitive, but it is also a lot of work to extract the data. Not an easy task at all.

Quality is his final factor, which is how good the business is, regardless of stock price. Many people use ROE, but I do not fully agree. I think that the EBITDA/EV is a much better metric to see how the business uses its resources.

On a side note, another part of Quality is actually the financial strength of the business. Again, this metric is regardless of stock price. A good strong business would be financially strong, and that means having plenty of current assets to meet current liabilities (current ratio, but I like quick ratio better), having a decent amount of cash for a rainy day (as opposed to distributing it all back to shareholders and operating too close for comfort) and to finally to have a low amount of debt.

As Peter Lynch famously said, "It is very hard to go bankrupt if you don't have any debt. It's tricky, some people try to approach that, but that's a real achievement."

However, I think my main takeaway from reading Patrick's articles is that just plain vanilla diversification within an equity index is not going to reap you the best results. Although you definitely get the good companies, you also get the bad ones as well. This is not to say that index investing is bad, but I am trying to say that it is possible to outperform an entire index, if you are able to pick the better performers and cut loose the weaker ones. But how would you know which ones?

Certain factors have proven to be statistically relevant is determining the medium and long term prospects of a business (and therefore it's stock price), so just having simple screens to weed out bad businesses or even good businesses that are ridiculously priced, you will be left with a decent batch of companies that leave you with plenty of upside and capped downside.

This is by no means a guarantee that the cheaper, investor-centric, high quality companies with low risks of bankruptcy will perform better than a mixed basket of stocks. The stock market is, of course, irrational.

I shall end of by including a graph that Patrick made. Screening dramatically improves performance by a huge factor, especially when the differences compounds over many years.


If you merely invest in stocks by just simply looking at P/E ratios and their dividend yield, maybe now would be a good time to rethink your framework and strategy?

No comments:

Post a Comment

Observe the house rules.