[Week 32 of 2026] Lists and Labels

Welcome back to Price and Prejudice with a few musings from Week 32 of 2026.

Prescriptions Without Doses

This post from Cove Street Capital features a list titled "Twenty Investment Lessons of 2008," which the author attributes to a Seth Klarman interview. The list is quite comprehensive – #1 is that things that have never happened before are bound to occur with some regularity, and #2 is that when excesses like lax lending standards persist, people are lulled into a false sense of security. #9, in full, is "You must buy on the way down."

While the list is informative and serves as a very useful mental model for investments, it is also one that's difficult to implement in practice. "Buy on the way down" is not a good rule – down from what, in what size, and for how long? The same issue with #2, where excesses persist "for some time" – anyone who knew the duration would not need the lesson. Of course, the generous reading is that these were never meant as decision rules at all. It's more like a constitution, not a trading manual, and I suppose constitutions are supposed to be vague, even for investments.

Twins Separated

This WSJ column reports that index funds claiming the same exposure are producing wildly different returns this year. The iShares Russell 1000 Growth ETF is up 4.6% year to date while Invesco S&P 500 Pure Growth is up 24.3%. Among leading large value ETFs the spread between best and worst is eleven percentage points. Apparently, most of this is due to two stocks: Micron and Sandisk, both memory chip makers, who are both up enormously this year at 188.6% and 411.8% through July. The interesting bit is that it sat in the Russell 1000 Value index until the end of June, because Russell's definition of value leans on price to book and companies that own fabs and equipment score well on book value.

The writer's recommendation is to stop slicing and own the whole market, and he is probably right for the marginal individual investors. But for many, they cannot take that advice. Most of the current infrastructure – institutional policy portfolios, target date glidepaths, and adviser model portfolios – are all written in style box language. This means that in addition to the decision of whether or not to own value, you also have to examine the definition of value to buy.

This second decision is one almost nobody had to think about so far. Usually we do not pay much attention to the provider's definition of what you are buying, so it gets settled by default, usually by whoever built the model portfolio three years ago and picked a fund off a list. When a choice goes unframed like that, people fall back on the one number that is always in front of them, which is the trailing return. So the flows this year will go to whichever provider happened to be holding Micron, which is not a view about value at all.

Podcasts

• I found this episode interviewing Mike Kelly quite interesting. Apparently in his thirties, he stopped trying to become a great investor, recognizing that getting to the top decile requires a lot of luck stacked on top of the skill, and that it is an enormously crowded pond. Instead, he chose to build and run asset management companies, by contrast, was a "pond nobody seemed to be fishing in."