Rules For Selling Stocks

“Selling is well and truly the dark continent of investing.” Here the author attempts to explain how a broker identifies high conviction sells.

A year ago when I published “Gurus of Chaos”, a book on the seven fund managers whose lessons have helped me find my way through the Indian market, I noticed that most investment gurus tend to find selling much harder than buying. As Sanjoy Bhattacharya, the founding CIO of HDFC Mutual Fund, says in Chapter 1 of my book, “Selling is well and truly the dark continent of investing.” So, given that there are many more books on buying than on selling stocks, I thought I would take a stab at how I as a broker identify high conviction sells.

Let me begin by highlighting two of the weaker “selling rules”. The first is around cyclical stocks and the economic cycle – common wisdom says that if the economic cycle is turning up, you should buy stocks from cyclical sectors such as Banking, Auto and Industrials (and vice versa). There is some logic in this: the correlation of stock prices in cyclical sectors with GDP growth is around 45-50%. The problem, however, is that the champion cyclical stocks such as HDFC Bank and Maruti demonstrate much lower levels of cyclicality (around 25-30%). Hence to buy or sell the top quality cyclical names on the back of the macro-cycle might not make that much sense.

The second weak rule that often gets used is to sell a company’s shares after it has surprised negatively in its results (and vice versa). Leaving aside the obvious issue that the stock will react very quickly to results and hence the incremental upside left could be modest at best, there is another problem with this rule – most companies tend to have good and bad results in an alternating fashion as corporate profitability is usually a mean-reverting series. Hence, an investor who responds to results could find himself buying and then selling the same name in six-month cycles that are unlikely to make him a successful investor. That being said, there is some logic in selling a company’s shares if the results are constantly below expectations.

Let’s now turn to the more effective selling rules. Firstly, it makes sense to sell a stock when you can see that the competitive moat around that company is eroding. Most investors tend to be slow in terms of coming to terms with the fact that a well-established company is gradually losing its moat due to changes in the wider world. For example, for more than a decade after Maruti entered the Indian market, Hindustan Motors, the manufacturer of the now obsolete Ambassador car, was a Sensex company.

Even after the opening up of the Indian economy in 1991, investors did not figure out that this company’s moat was going to be eroded by economic liberalization. Last year – a full 30 years after Maruti’s entry into India – Hindustan Motors shut down its manufacturing plant. I reckon investors face a similar risk in Indian banking, housing finance and NBFC stocks given that the RBI is now whacking our new banking licenses by the dozen. These new banks could erode the incumbent lenders (I am referring to incumbent banks and non-banks) on both sides of their balance sheet.

Secondly, it makes sense to sell I believe when a company’s capital allocation is deteriorating. ROCE, as a measure of the effectiveness of a company’s capital allocation, is the most accurate predictor of stock prices and anybody who can predict a company’s ROCE evolution will become a very successful investor. Great companies start sliding when they can no longer figure out how to invest their next dollar of free cashflow at a higher ROCE than their current ROCE. Bharti and Tata Steel are the best examples of titans who have suffered on this front. Asian Paints, on the other hand, has been a champion of the capital allocation game over the past 30 years.

Finally, the most effective selling rule I believe, is centred on accounting quality. When a company starts cooking its books, it is time for minority shareholders to head for the door. Every year in the month of December, my colleagues and I crunch through the last six years of annual reports of the BSE 500 companies. (We wait until December because most of the dodgy companies publish their annual reports in November.) We then rank companies in each sector on their accounting quality basis a battery of forensic accounting ratios such as Operating Cashflow/Operating Profit, Asset Turnover, Provisioning for doubtful debtors, etc. What we find every year is that the bottom 40-50% of Indian companies (i.e. companies with weak accounting) do not provide any shareholder returns over the medium-long run.

At one level this isn’t surprising – a company whose promoter is shortchanging minority shareholders is unlikely to be generating shareholder returns over the medium-long run. However, what is interesting is that large, well-researched companies (with between 20 to 50 brokerage houses covering the name) also tend to have glaring accounting issues which the stock market sometimes takes years to unearth. So much for the efficient market hypothesis!