October 15, 2005

Trying to develop a mental model to value cyclical / Commodity businesses

I have always had a mental block against low return commodity / cyclical businesses (highly influenced by warren buffett’s writings).

Other than the fact that investing in such businesses requires being tuned to the business cycle and overall requires more work in tracking such businesses, I have found it difficult to value such businesses. It would be naïve to use PE or such simplistic ratios because these ratios would mislead you completely. During the peak of cycle, due to operating leverage the earnings shoot up and the PE drops to single digit, making the stock appear cheap. The reverse happens when the business cycle turns.

Developing a DCF model also has been difficult, because I have found it difficult to predict future free cash flows for such companies (I assume that would require having a reasonable grasp of the business cycle in terms of demand supply picture, inventory levels, pricing etc).

In contrast businesses like FMCG, paints, pharma have good returns, low or non-existent cyclicality. I have found such businesses easy to understand, project for some years out and value them. In addition, for the past few years such business were available at throwaway prices. As warren buffet says (I like to quote him a lot), “degree of difficulty does not count in investing, being right counts more. Better to invest in a simple business with a single knowable variable, than a complex business which have multiple complex factors driving it” (I have paraphrased from memory).

So why this change of heart. For one, it is an intellectual challenge. I would like to understand and value such businesses, even if do not invest in them. In addition, it increases the investible universe for me.

It could a long time for me to confidently value such businesses. The ‘business analysis’ excel is an effort in that direction. I also have a detailed valuation file (which I will post shortly), which I use to value an individual company. Hopefully towards the end of this effort I should be able to use it to value a commodity/ cyclical business.

No comments: