Hi Thomas,
I would have replied earlier, but a string of interruptions has deflected me for a while. More good questions, a few thoughts to chew on. Hardly complete answers. Time is precious at the moment.
One is what is the basis of the metric one will use. By this statement you are pointing at a family of metrics. Which one should we pick and why?
As Einstein said "All models are wrong, some are merely useful". I believe the answer to which one is "several". Because all will be wrong, but not all to the same degree, and not always in the same direction. Triangulation works in the woods with a compass. Works in the market as well. At least, IMHO.
Because the thing that makes it hard is choice. If T-Bills are getting 2% then a company with profit of X is worth way more than when T-bills are getting 12%.
So one must triangulate not only using several models, but on the results they generate for a bunch of different datapoints. Kind of like a GPS system where the location of none of the satellites are known to begin with. Reminiscent of the work of Keppler and Copernicus.
Another is what do we do with this metric once we obtain it? Is it merely a rude screen that keeps us from doing something way out of line or does it provide some kind of buy and sell signals? How do we arrive at the values of those signals? What kind of confidence should we have in them? Are there other factors we should consider in combination with these signals?
Yes a rude screen - always. Never alone. Buying great economics with lousy management is also a recipe for disaster. Buying the rights to betamax would have been a bad idea too. It's all about the mix. Each element should have "veto" rather than "voting" rights. No sense buying a great technology with great management that will never pay back. For example.
Buy signals? Sell Signals? Absolutely. But always as a range rather than some magic number. Like a thermostat set at 72 degrees won't turn on at 71.999 and turn off at 72.001 or it would hunt all the time. The width of the comfort zone depends on market factors and risk tolerance. This is where it gets kinda personal. If one is willing to take lots of risk one might set the sell band very far away from midpoint. Etc.
Confidence? One should have absolute confidence in whatever scheme one chooses. And confidence that it will be wrong. At least to some degree. The point isn't to eliminate risk by engineered precision, but to contain risk within tolerable levels and attract commensurate rewards.
Plus one must have both the humility to recognize that it's ones' own errors that it generates and the inquisitiveness necessary to learn and refine as one snatches the inevitable "learning opportunities". First thing I had to get used to is the concept of being wrong all the time. Hopefully progressively less so as time goes by. This is where Mucho Maas' concept of portfolio investment is very very handy, if only as a risk mitigation eggs in many baskets kind of sense.
Other factors? Yes. We are discussing risk management and value harvesting rather than just absolute crystal ball gazing. So there is company specific risk (e.g. LU vs Cisco) then there is sector specific risk (e.g. Telecosm implosion), then market risk (bubble) then macro risk (e.g. WTC). All are playing together to a greater or lesser extent. It is somewhat amazing to see them all line up in our lifetime.
And then you asked two versions of: How well does share price tie into your preferred metric over time? How extended are the periods where the two are out of whack?
Share price "ties in" kind of like room temperature ties into a thermostat. Depending on what it's at the furnace may or may not turn on. Loosely speaking.
How extended are the periods where the two are out of whack? Always. I have never seen shares priced "at" what I consider to be economic value. Generally they are all over the map, both above and below. One would expect that the price should be below a target computed absent risk, and that the difference would represent profit opportunity commensurate with the level of risk. I am a conservative person so I tend to find my modelling gives me targets that are higher than the market's asymptotes. Which of course I factor into my decision making process. Each person will (should) introduce their own personal bias. It's rather hard to avoid. IMHO. It's also hard to recognize.
But one doesn't have to look much further than the nearest gas station to see that "value" and "risk" are not the only factors at play in pricing decisions. There is usually something about the elasticity of price with respect to supply and/or demand shifts. Which in our boomer dominated market I have found to be incredibly long lasting. All of a sudden "Optical" is "the thing" and everything in the sector gets bid up. Then it's Wireless, or maybe Storage.
I further suggest that there is two kinds of "out of whack" to think about. There's the sudden "way-out-of-whack" kind of being bid up factor ten or trashed by factor three kind of out of whack. Where it makes sense to sell or stay well away from or buy as the case may be.
Then there's the out of whack by 20%-30% or so. In which case "no decision" is prudent. That is, if in, stay in. If out, stay out.
And then finally a note that the whole thing changes with time. If one were to compute Cisco's apparent value thirty years ago one would get a different answer than today (yes, I am being extreme and it's only for illustration). The same holds true to a lesser extent as the time periods get shorter. I like to refine views quarterly. You may see reference to my posts saying "I have opinion X now, but I am looking for Y before revising my view". Which isn't being vague for the sake of being shifty. It is merely recognition that exogenous events can have sweeping effects.
Too much, I'm sure.
John. |
