the heavy half and portfolio management

Consumer products marketers often talk about the “heavy half.” The idea is that a relatively small subset of a product’s buyers use the product often enough that they generate most of the sales. Say, 10% of the Bud Lite users account for 60% of the cases sold. The Pareto Principle, the idea that 20% of the customers create 80% of the sales, is another way of framing the same idea.

There’s something like the heavy half in equity portfolio management, especially for growth-style managers, who typically hold much more concentrated portfolios–maybe 50 stocks, maybe fewer–than their value counterparts, who can have a couple of hundred names. In all likelihood, three or four of the stocks in a growth portfolio will end up creating most of its outperformance.

The other 45+ are there for two related reasons:

–no one knows, or at least I don’t, when an idea will catch fire in the overall market. If I’m lucky I have a pool of ten solid ideas that I think could do very well. Experience tells me that only two or three names will perform in the way I’m expecting, but I really don’t know which ones they’ll be. So I hedge.

–the rest of the portfolio is there for defense, to force me to be aware of what’s happening in sectors I see less potential in and to mitigate the damage if the areas I’ve taken money from to build up my overweights begin to perform better than I think they should. Yes, I expect to make money in these areas, but it’s more important to me that they not punch a huge hole in the bottom of my boat.

Not everyone follows this approach. Take the ARKK portfolio of billionaire Cathie Wood. Its results over the past five years are as follows:

Tesla +1300%


S&P 500 +55%

ARKK -7%.

Ms. Wood understood far earlier than the consensus (me included) the profound changes in the automotive industry that TSLA would set in motion. If she had become a sell-side auto analyst, she might well be the most famous securities analyst on the planet.

If we assume that she maintained an average 5% position in TSLA in her ARKK fund over the past five years (I’ve plucked a figure out of the air that seems right to me, but I make no claim that this is correct; the TSLA position is 11.8% now). If so, TSLA would have contributed about +65% to overall portfolio performance. This would imply that the non-TSLA portion of the portfolio lost something like 60% of its value over the last-half decade (even after adjusting for distributions amounting to about 10% of nav over the period). This is in a NASDAQ market that was up by 73%.

If the average TSLA position were 2%, its contribution to overall results would be +26% and the rest of the portfolio would have lost about a quarter of its value.

I’ve been sitting here for a while, thinking to myself that, given the apparent magnitude of the underperformance, the last four paragraphs can’t possibly be correct. I got the performance figures from Yahoo Finance, however, and the distribution amounts from the ARK website. I did add the 10% to ARKK performance rather than doing a time-weighted calculation, but that shouldn’t have made much difference–and the NASDAQ and S&P results remain capital changes, not total returns, giving ARKK an edge.

I guess it’s possible Ms. Wood traded TSLA badly or that she abandoned what I think has been her best idea for part of the time. Still, this is an example, more extreme than I’d realized, that defense counts for a lot.

the ARM IPO next week

TSMC and ARM are the two most important chip companies from late last century (or 20-some-odd years ago, if we want less drama). TSMC allowed talented chip designers to break away from the lumbering integrated giants of the day–INTC or TXN, for instance–by offering them access to state-of-the-art manufacturing facilities. So they didn’t have to come up with billions of dollars to build their own fabs. ARM offered them standardized software tools for rent, so designers could concentrate on innovation rather than spend a ton of time creating their own hammers and saws.

Softbank, a Japanese investment company controlled by Masayoshi Son, which after a brilliant start has had a tough 21-century so far, owns ARM. It attempted to sell ARM to NVDA for $40 billion in 2020, but was shot down by anti-trust concerns. So it has decided on an IPO instead, with ARM making its stock market debut on September 13th.

I haven’t read the preliminary prospectus, which is available on the SEC’s Edgar site (ARM is going to be the ticker symbol). The offering price, just under a tenth of the company being sold, is being set as $47-$51 per share. This would value ARM as a whole at around $50 billion.

What I find interesting, and the reason I’m writing this, is that I’ve already gotten a blast email from Fidelity offering the possibility of getting an allocation of shares.

I don’t read this as a sign that demand is red-hot. It’s hard to know if the issue is the company/price or whether the apparent hodge-podge of US, UK and Japanese bankers is closer to the Mets/Yankees than the Braves/Orioles.

entering September

April may be the cruelest month, but for investors September (really, the four or five weeks from mid-September to mid-October) is by a considerable margin the worst month for US stock performance.

The reason is pretty simple. Mutual funds are required by law to distribute basically all their realized capital gains and net dividend income to shareholders. That’s because fund holders have to pay income tax on these distributions and Washington wants to collect income tax as fast as possible. So the government pushes funds to end the fiscal year on Halloween, something that every one I’m aware of does. That gives a fund time to close the books and still pay shareholders before the current tax year ends.

Oddly, to my mind anyway, fund holders like to get distributions and regard them as a mark of fund success. So the approach of the fiscal year end has become the industry’s trigger for managers to realize capital gains. And that turns into the occasion for making a general tune-up of the portfolio. Funds trim outsized winners to create funds for distribution to shareholders. And they also tend to use this tax-selling time to toss clunkers overboard and realize capital losses. In both cases, they sell.

This selling usually begins around the second week in September and dries up in the middle of October, giving trades ample time to settle before the fiscal year end.

Last year during this period, selling was unusually severe, with NASDAQ losing about 15% and the S&P 500 a bit less. This decline also came after a ~5% fall from mid-August highs. But the end of selling also marked an important market bottom.

My hunch is that this year’s experience will be relatively tame, both in comparison with 2022 and in absolute terms. I’m not willing to bet the farm on this thought, but I’m also content not to do any selling based on worries about a repeat of the 2022 losses in 2023.

My biggest concern is that trading bots will mindlessly try to replicate last year’s trading. It will be interesting to see if they do, and, if so, how successful they are.

Two thoughts for you and me:

–there’s nothing to prevent us from doing a portfolio overhaul of our own, something we should all do regularly, and

–if bots run wild, we may have a great chance to buy in a few weeks.

Dollar General (DG)

DG reported disappointing quarterly results overnight and is down by almost 17% in pre-market trading. Two things:

–it will be interesting to see how the stock fares once the market is open. I read the sharp drop as bots being more aggressive and lack of buy-and-hold participation in the aftermarket. I have no real thoughts on how trading will go, just that a bounceback would be evidence that trading against bots may not be as foolish as might seem …and the lack of one, or a further decline, will be equally instructuve

–I see the pecking order of traditional retail, moving from the most affluent customer base to the least, as:

luxury goods

department stores (maybe an anachronism)

Target (TGT)

Walmart (WMT)

dollar stores.

In good times, customers shift upmarket from where they usually shop. In bad times, they shift down.

As I read the DG news, the company is facing two issues: it has too much inventory (a la TGT) and customers who shifted down during the pandemic are beginning to return back up. It may well be able to keep some of this higher-end traffic and this long-term plus is being obscured by the inventory issue. Still, the message for the economy is that buyers are giving an all-clear signal. The chief beneficiary should be WMT.

WMT, in turn, will be hoping to keep customers who have shifted down from TGT.

If I were still an institutional portfolio manager, and if I were interested in US mainstream retail (which, I think, we all should be) I would have a more-or-less permanent position that would consist of TGT + WMT. In a good economy, I’d be much heavier in TGT; in a bad economy, I’d be heavier in WMT, with maybe a smaller position in dollar stores. Right now, I’d be shifting out of my more defensive, WMT-heavy stance.

As an individual investor, I’m much more interested in IT than retail. I’ve had a substantial position in TGT for years. It has been a bad mistake over the past year not to have shifted into WMT. But I find myself just starting to add to TGT again.

Another thought: I think there’s a big question mark over luxury retail. A significant chunk of this sector’s revenues (20%?) and a larger part of profits (1/3?) come from China. Not a good place to be right now, and potentially a source of negative earnings surprise over the coming few quarters. Another potential plus for other retail stocks.