March 3, 2024 —
February was an interesting month in financial markets. Nvidia stock continued its incredible run on the back of a huge earnings beat. Tech did very well (again). Talk of tech bubbles has become pervasive, though serious people have pushed back against the tech doomsayers. (Our view is that tech stocks are not in bubble territory, a point we made last June.) A hotter-than-expected inflation print led to a rise in Treasury rates, and a push back of the anticipated start date of Fed easing (from May to July of 2024 according to Fed funds futures markets). And, perhaps spurred by investor inflows into newly approved spot ETFs, bitcoin rallied 45% in February. Indeed, leap-year February 2024 did not disappoint.
Alongside tech and crypto, the stock market had a good February too, led by momentum stocks (i.e., past winners) and technology. Mid- and small-caps did well, and even financials managed a 4%+ gain in the month. The worst performers were all things fixed income, especially the high duration parts of the space, like long-dated U.S. Treasuries. Commodities, gold, and utilities managed to eke out small gains.
For our part, QuantStreet’s three risk-targeted separately managed accounts did well, and again outperformed our mutual fund asset allocation benchmarks. You can read more about our performance here.
Research
In going through some client tax returns for 2023, we noticed an interesting phenomenon: Many accounts, which were up handily in 2023, showed net realized losses on the year. How did our portfolios turn that neat trick? The key lies in the trend part of our trading signal (the other part is our machine learning model forecast). When an asset class starts falling in value, its trend deteriorates, which makes it look less attractive to our portfolio allocation engine. So our process naturally kicks out past losers, which can lead, in some cases, to portfolios that are up year-over-year but which still show capital losses for tax purposes. We are intrigued by this observation, and are in the process of better understanding this aspect of our dynamic trading strategy. We’ll report back on this in future letters.
One way in which our portfolios differ from those of our competitor asset allocation mutual funds is our relative underexposure to international assets. This is despite our being keenly aware of the benefits of international diversification over time periods excluding the last 20-30 years. The main reason for our international under-allocation is that our combined machine learning-trend signals do not favor international or emerging market investments at the moment. We very much believe in a process-driven investing approach (with human oversight), and should our models change their minds, we will surely allocate more internationally. But given our current underinvestment, we have spent a bit of time thinking about why U.S. stocks have outperformed so much over the last several decades. The net result of this effort is a recent piece in which we show that international markets look relatively less attractive than the U.S. along measures (put out by the CATO Institute) of human freedom, property rights, and rule of law. As it turns out, the freedom gap between the U.S. and many other parts of the world is growing, which we suspect bodes well for continued U.S. outperformance.
Finally, we admit to a pet peeve: People often put up charts comparing the stock performance of a currently high-flying stock to that of a past high-flyer that subsequently went on to have terrible returns (like Enron). The implication of such analyses is that, if it happened to Enron, and if XYZ LLM Bitcoin Inc. has a price chart that looks like Enron’s, then surely XYZ will meet a similarly inglorious end. If such arguments also strike you as fishy, you may enjoy reading our piece entitled “What Does Enron Teach Us About Nvidia?” (Quick preview: Nothing.)
Portfolio updates
For the month ahead, we largely retain our prior positioning, which was underweight small-caps, midcaps, and all things international, with overweights in technology and healthcare. See last month’s client letter for our reasoning. This month we slightly reduce our U.S. Treasury exposure and allocate a small part of the portfolio to gold. Why? First, our machine learning forecasting model seems to do a reasonably good job of forecasting gold returns, for reasons that appear to have to do with the cost of carrying a gold position and investor concerns about inflation. The model currently likes gold, and the last one-year price trend in gold also looks good. On top of this, gold is a great diversifier for our portfolio because it has a low correlation with most of our risk assets. Finally, our analysis shows that gold typically does well during Fed easing cycles, and according to the Fed funds futures markets, such an easing cycle is likely to commence starting in July. We will be putting out a detailed piece on our gold analysis soon. If you’d like to receive it, please sign up for our Substack page (if you are not already signed up).
Working with QuantStreet
QuantStreet offers wealth planning, separately managed accounts, model portfolios, and consulting services to our clients. If you are an existing client or if you are thinking about working with us, we’d love to hear from you. Please reach out to us at hello@quantstreetcapital.com.