The BuyChart of the day: TSMC Expectations
We had been holders of TSMC for nearly 5 years. A couple of months ago, we started unwinding our position – not because we don’t believe in its competitive advantage or the (massive) thematic tailwind that it is both fueling and riding – but because we couldn’t justify the stock price anymore. That view was based on TSMC’s Buycast, which got worse with last week’s earnings call. Now, we believe that that negative year over year revenue growth in Q2 is an aberration that will soon be rectified. But it’s still hard for us to digest this chart below from the Buycaster. Basically, we’re not in the mood to buy TSM now, but we will always look for another rational entry point again. Not yet. Not yet.
The BuyTheme of the day: AI & Big Data
A couple of weeks ago, we had published a chart depicting Buycaster Ratings of a pretty decent set of AI & Big Data leaders in the semiconductor space. We still believe the semiconductor industry is the best way to play AI & Big Data. But as we’ve seen recently, it goes through periods of irrational exuberance. Here is an update of that chart. Some stocks haven’t lost their sheen. TSMC and Tokyo Electron the biggest movers – albeit in opposite directions.
Macro Dose: Valuation Uncertainty Principle
Nick Maggiulli (of Dollars & Data) raises a good point: All valuation models are flawed. Let’s accept it. Let’s stop this business of predicting. There is no deterministic formula for success in investing. There is no one magic metric that will tell us whether the market is overpriced or underpriced. Investing is a probabilistic exercise. As Maggiulli points out, all those metrics that pretend to predict the direction of the market (or stock) assume some sort of “mean reversion”. That’s the idea that a metric will revert to some sort of “natural equilibrium” in due course. For stocks, the term for this “natural equilibrium” is Intrinsic Value. There is no Intrinsic Value.
A long time ago, we had compared the concept of Intrinsic Value to the Heisenberg Uncertainty Principle. Intrinsic Value is unknowable because the factors required to calculate an accurate number are too varied and complex, even for AI (as of now); and if it is knowable, the amount of time it would take to calculate it would render the calculation obsolete. "Intrinsic Value" will have moved on.
Have you seen Oppenheimer yet??