On Grahamian & Fisherian Framework

(Following is a transcript of a mail, which I had written sometime back, to a fellow value investor cum colleague of mine on the scope of Grahamian and Fisherian Investment framework)

Date: Mon, 18 Feb 2008 05:24:29 -0800 (PST)
From: Arpit Ranka
Subject: Discussing Graham & Fisher

Following our discussion a couple of weeks ago, I spent sometime reading Fisher's 'Contrarian Investor Sleeps Well.' And I have to say that it helped put things in perspective, in quite an unexpected and interesting way... :-)

Let me put forward for your consideration and comments, the key lessons that I have realized, pertaining to the whole discussion on the scope of Graham and Fisher Investment methods.

1) Understanding Stock Returns:

One thing I realized is that, irrespective of whether you subscribe to Graham or Fisher, the importance of PE expansion in helping you derive above average returns cannot be underestimated. For ex:

As a Grahamian investor, looking for low PE stocks representing decent businesses, majority of the gains are earned in the form of PE expansion and not in terms of EPS growth. The prospect of EPS growth ensures that while you wait, you are being paid to wait with the commensurate increase in intrinsic value, and also equally importantly, in some cases, it might also act as a catalyst. It would not be far off to say that it is akin to trading, where you are betting on a positive reversal of appraisal of a business by the market participants based on fundamentals.

As a Fisher proponent, if one is looking to buy great businesses, which are few in number at around, say, 15-20 x earnings for a business with a prospect of growing earnings at 15% and generating/maintaining high returns on capital employed - the chances of generating returns using PE expansion over the long run are minimal - unless one is betting on speculative increase, just like in Grahamian framework, in positive reversal of appraisal of such businesses - and one can only expect to earn the returns that reflect the performance of the corporation in the long run.

For ex: if I buy into something at 20 x earnings, and expect the earnings to grow at 15% over the next five years, and have reason to believe that the quality of business would be maintained, five years hence, to ensure that the earnings would be appraised at 20 x earnings then, I would get a 15% return.


2) Non-Linear relationship between Known and Unknown; Expected and Unexpected:

The only problem is that to execute Fisher's investment methodology and be successful at it requires a couple of things - 1) availability of businesses where future can be predicted, with considerable certainty, and those predictions can be comfortably backed by capital today, and 2) ability on the part of the investor to have the competence to judge those sort of businesses, which no matter how hard one tries would require superior understanding and experience, of the whole business landscape, to go along with the psychological requirements of being patient along the way.

This brings us to the discussion of non-linear relationship between known and unknown on the part of the investor; and what can be expected and not expected from the business performance going forward. The things that can be known and expected constitute not much of what that cannot be known and expected, yet in the Fisher approach we are required to back with more capital and time, in the form of holding periods, than in the Grahamian approach, on what can be known and expected. That is, the margin of safety - in terms of betting on your own understanding of a situation - is put to test more vigorously in Fisher's approach than in Graham's.

3) Re-investment Risk, Transaction Costs & Taxes:

I think a couple of factors that add considerably to the Fisher approach, when rightly implemented, and not to Graham's approach are 1) Re-investment risk and 2) Transaction Costs & taxes. The very fact that Graham's approach leads one close to trading like approach - selling on the expansion of PE ratios and looking to invest in other cheap stuff exposes one to re-investment risk (a big concern for people operating with huge amounts of capital and, resultant, smaller universe of stocks to choose from) and higher transaction cost and taxes over the long run. But barring the first constraint of large pool of capital to invest, I think transaction cost and taxes are sufficiently compensated by the lack of precision required in generating above average returns from Grahamian approach, unlike Fisher's approach, which without doubt is much difficult to successfully implement.

Considering the above and PE expansion criteria, I think that Fisher approach would be a better way to deploy capital, particularly when businesses of great quality can be bought at throw away prices, like Mr. Buffett did by buying into great businesses at 1973-74 and 1987 period. That way we would not be paying much for future growth and the growth, at high returns, which can be considered probable, would provide icing on the cake. Towards this end, I think Mr. Buffett has successfully managed to seam the best from both the schools and his results speak for his brilliance.... Hats of to Fisher and his successful disciples!

-----------------

Considering the overall scope and limitations, particularly relating to human competence and conditions requisite, for successful implementation of these strategies - I have come out with a lot more appreciation and reinforced vigor towards Graham's framework of sticking with buying good businesses at cheap prices...

Moreover, if any transition is to happen, I think it has to be a natural evolution based on increased understanding of great majority of businesses operating in the corporate landscape leading to a conviction about being capable of seeing the prospects of such businesses far, far into the future and more importantly, exceedingly depressed market/security specific levels to start thinking of taking a very long term view (10 years plus) on an opportunity.

It would be great to hear your thoughts on this, if and when time permits.

Thanks & rgds,
Arpit Ranka