No. of Recommendations: 27
What do you think should be the second step? That's easy.
Find the overoptimism in your assumptions, squeeze that out, repeat : )
Long ago I based my models for IBM on the notion that they would, sooner or later, make $20 a share, and the biggest unknown was the time frame.
My forecast for the likely time frame was more conservative than all the investment houses and analysts, but that doesn't help if they never manage it.
Since then EPS has fallen at about a buck per share per year per year--oops. The lesson was that there was way too much optimism about their business prospects, and way too much optimism about my understanding of why and how much it was predictable. I was in good company, but that is little consolation. It was my second biggest loss ever.
Part of the predictability is running through a list of obvious flaws, and ruling out any candidate that might have one.
Too much leverage, too critical on a legal situation that might reasonably be expected to change, signs that management might possibly be less than 100% trustworthy, a superficially tempting winner in an industry known for its horrible economics.
Make a list of 20 possible big red flags--it's not hard. Then use the list. That's hard. Remember the Nike slogan.
For some firms, you'll want to model sales and margin quite separately.
Some firms have relatively trending sales but oscillating margins, so extrapolating sales and making assumptions about normalized margins can help improve things.
For example, the table in this post on Alphabet can be a useful template
https://www.shrewdm.com/MB?pid=687699769For a few firms of course, future earnings aren't the best proxy for value. Berkshire is the obvious example here.
Fun fact:
If you use peak-to-date book-per-share as your metric for BRK, then the rate of change of that metric is in effect the company's ROE.
For all four year rolling intervals (quarterly snapshots) for stretches ending March 2013 through ending June 2022, the rate of change was inflation + 9.9%.
The remarkable thing was the extraordinary steadiness: the standard deviation was less than 1%. Extremes were inflation + 8.6% to inflation + 11.7%.
Even if you don't use peak-to-date for book which has a smoothing effect, the standard deviation of those rolling-four-year results was only 1.3%.
The latest figure is inflation + 7.6%...still in the old range if you're looking at nominal numbers, but the after-inflation number is flagging.
(some part of that is the fact that working assets on the books like aircraft or furniture stores don't get marked up in value with inflation, but working assets aren't really really a big part of the pie)
The two observations from this:
(1) The steadiness really is remarkable. For now, assuming inflation plus 7-8%/year for the next 5-10 years doesn't seem to be aggressive. I count on 7, kind of expect 8, hope for more.
(2) The absolute number isn't that great. But I'm cool with that. I'll make it up on volume.
Also, do you think having deep knowledge of the company and industry is necessary or critical to using an approach like this to select strong investment candidates, or do you think that's more icing on the cake? I think it's necessary for only one big reason:
You have to understand a company's economics, how and why they make their money, to be able to assess how likely it is to change for the better or for the worse.
You can't stamp out the optimism hidden in your projections if you can't see it.
Similarly, a relatively deep understanding is needed to have a feel for what big unexpected things might go wrong.
For example, Apple's numbers alone won't tell you that the big risk (in my eyes) is the Chinese Communist Party because of the extraordinarily large bet on China operations and sales.
I used to figure I had a decent handle on the "pretty darned sure" outlook for maybe 20 companies, and real confidence in maybe 10.
It has taken my many many years of work and study since then to improve that down to about five now. I have much more knowledge about many more businesses, but the dominating factor is improved knowledge and appreciation of how little I know and how unpredictable the world is.
Of course, that doesn't mean you have to be the deepest expert in every industry. Some industries are inherently simpler than others, so extreme depth isn't as needed in those. That's where I hunt!
For example, other than a bit of inflation on the sticker, dollar stores haven't really changed since the days of Woolworths five-and-dime stores.
Some industries are full of unpredictable firms, but the industry as a whole (a scatter-shot bet) is relatively predictable.
For example, somebody who owns small pieces in all medical device and medical supply firms will probably do pretty well. Individually they are hard to follow, but as a group they have excellent economics and long run returns--perhaps that's likely to continue?
Average return last 10 years among the two relevant industry categories at Value Line is 13.2%/year, which is 3.9%/year better than the average firm in their database.
A similar thing was once noted by Mr Buffett about drug firms.
My biggest insight:
Never look for the companies that have the highest likely returns--too many people looking there, the uncertainty is high, and on average people overpay for growth so they're rarely undervalued.
Look instead for the assets that you (justifiably) have the highest confidence in their futures, even if they are slow growers, or even cash cows or cigar butts.
Whether it's stocks or real estate or Beanie Babies, if you know what something is worth better than the average player, it's not hard to find a way to make money from that knowledge.
But you still want high returns? Go to the dark side. FIRST find something you have supreme confidence in your ability to predict, even if it's a slow grower. THEN add a pinch of leverage. It doesn't take much at all to turn an uninteresting rate of return into a very interesting one.
Just never fund the leverage with a loan that can be called.
Jim
