On The Intelligent Investor: Part 1 — Importance of Industry Selection
Warren Buffet is not shy when it comes to giving away his investment strategy. In fact, he attributes 85% of his investment strategy to a single book, The Intelligent Investor, by Benjamin Graham. The Intelligent Investor was originally published in 1949 and revised most recently in 1973. Despite this fact, Buffet claims its wisdom is timeless and no investment book since has come close.
So I decided to read it and write about it in a multi-part series.
Reading the book inspired me to build a tool for screening investments, Stox. Stox is a stock screening platform that allows you to filter and analyze stocks using natural language. Check out more at https://stox.dev
You can check out the jupyter notebook for this post here.
Are you an intelligent investor?
Intelligent Investor doesn’t focus so much on analyzing securities, rather on investor principles and attitudes. It also takes more of a macro industry view:
It has long been the prevalent view that the art of successful investment lies first in the choice of those industries that are most likely to grow in the future and then in identifying the most promising companies in these industries
This strikes me as ultimately the correct approach. Companies in an industry tend to move together and that’s even more true today. If you were to have me guess the performance of a random stock over some period of time, the first thing I would ask is what industry to stock is in. If you told me the stock was in the airlines industry and the time period was Q2 2020, I’d have a pretty good idea of the stock’s performance relative to the market as a whole.
Importance of Industry Selection
I looked at all S&P 500 prices for 2019. I grouped the stocks to their sector, a broad category indicating the industry the company operates in. Here is the breakdown by sector.
Sector
Basic Materials 22
Communication Services 25
Consumer Cyclical 65
Consumer Defensive 35
Energy 25
Financial Services 69
Healthcare 62
Industrials 73
Real Estate 31
Technology 67
Utilities 28Don’t ask me why there’s 502 stocks.
Whenever dealing with time series information, it usually makes sense to look at the log differences. Logs essentially turn multiplication into division and are much easier to work with. Take for instance the prices below.
| price | price | log | log diff |
|:------------:|:------------------:|:------:|:--------------:|
| 2019-12-31 | 87.57 | 4.4724 | 0.0236 |
| 2019-12-30 | 85.53 | 4.4489 | 0.0019 |
| 2019-12-27 | 85.37 | 4.4470 | -0.0080 |
| 2019-12-26 | 86.06 | 4.4550 | 0.0063 |
| 2019-12-24 | 85.52 | 4.4488 | 0.0000 |
| | =ln(87.57 / 85.52) | | =sum(log diff) |
| | 0.023688207 | | 0.023688207 |If you take the logs of the prices and diff them (new - old), you can sum up all the diffs and get the log of the performance for the period. Then if you want you can take e^(sum log diff) to get the actual performance.
Let’s look at the median log diffs for 2019, broken down by sector.
| Sector | 2019 log return |
|:----------------------:|:---------------:|
| Technology | 35.16% |
| Industrials | 26.81% |
| Financial Services | 25.16% |
| Utilities | 24.61% |
| Healthcare | 23.91% |
| Consumer Cyclical | 23.15% |
| Consumer Defensive | 22.47% |
| Communication Services | 21.95% |
| Real Estate | 20.46% |
| Basic Materials | 17.36% |
| Energy | 3.42% |It was certainly a good year for all sectors, but there’s a big difference among some sectors. Let’s look at this over time.
The difference between the best and worst performing sectors is huge. I’m not saying that you could necessarily pick the best performing sector, but if you just excluded certain sectors or weighted them accordingly, you would affect your returns greatly.
Let’s look at how stocks did within sectors. I looked at the top 20% and bottom 20% returns. For instance, in 2019 the top 20% performing Industrial stock rose 39.75% (log return), while the bottom one rose only 10.88%.
The median stock in the Technology sector returns 35%. So if you had a crystal ball and invested in the top 20% performers of all stocks in Consumer Cyclical, Utilities, Financial Services, Communication Services or Energy, you did worse than a typical Technology stock. For everything apart from Industrials, you did about the same.
For 2018, this is not true. If you could pick the top 20% performers in Consumer Cyclical, the second biggest loser of the year, you’d get a great return of 11.86%, even better than if you had invested in the top 20% performers in the best performing industry (Utilities).
Final Thoughts
So how important is industry analysis? The differences of performance between industries and sectors is obviously very large. But depending on the year, the difference in the worst and best performers within an industry can be very large as well. But in regards to investment strategy, I agree with Graham that the industry weighting is the most important decision to make when investing in stocks.
The example of performance differences within industries looks at the top 20% of performers. This is an unrealistic measure. No one would be able to consistently pick the best 20% performers. A more reasonable expectation would be for an investor to look at the market and economy as a whole and decide which industries to over/under weight. Obviously you would want to over/under weight the individual stocks within the industry as well, but the first step should be industry selection, and if you stop there, that’s fine as well.
On the next post I’ll discuss avoiding the emotional aspect of investing.
