INSIGHTS
March 2026
Investment Notes and Insights
February 2026 ended with both a bang and a whimper, to twist a phrase from T.S. Eliot. The “bang” was the attack on Iran; the “whimper” was the aggregate loss in the U.S. stock market. The S&P 500® Index fell -0.77% for the month, concentrated in growth stocks which lost -3.35% as measured by the Russell 1000 Growth Index. Investment grade bonds were largely unchanged for the month.
More happily, all Advocacy Wealth portfolios generated positive gross returns when the investments were held for the entire month. All but one of the models had positive returns net of fee. All models rose comfortably above their annual target ranges net of fee year to date, though the sampled term is admittedly short.
How did we achieve these positive results? Diversification. If you have exposure to equities, you do indeed own some U.S. growth stocks within the funds that make up your equity allocation. But if you do, you also own many other types of stocks as well. You also likely own bonds as well as possibly some securities that have little correlation to the returns produced by either stocks or to bonds.
The big individual winning exposures for our clients so far this year are pretty much the same as for 2025:
This place may offer an abundance of wealth and possible solutions to the problems you brought with you.
In a forest, even a very dense one, there are many entry points. Here, are there paths? Are there hostiles or friendlies, or both?
Or do you bushwhack, and blaze your own trail to discover what can be uncovered in this new land?
Welcome to the shores of Artificial Intelligence.
Artificial Intelligence (“AI”) will likely transform our lives to at least an extent as did the Internet itself – upon which AI itself depends to be an actor. Its utility, in economic terms, will likely be a multiplier on the Internet itself.
Without electricity, AI does not exist. Without connectivity, AI does not exist. It is not as simple as unplugging that space heater that may burn your home down, but nor is it that different. So long as humankind controls both inputs, we should continue our existence. (Nod to HAL2000).
That is the “existential’ argument.
The internal argument is what do we do with Frankenstein, now that we have made him?
Dario Amodei, an OpenAi deserter who founded Anthropic with his sister and four others, worries about the jobs AI can do better, faster. The open question here is : cheaper? Sunk costs on at least $1 trillion in data centers has to be factored in. Yes, there is a reverse “J curve” possible: costs could plummet with wide adoption. But, as Davio worries, what is the social cost?
As commentators other than us have pointed out, wealth accrues to the economic beneficiaries of new technologies. Those beneficiaries are often not the inventor or the investors who invested in the inventor’s idea to bring it to life. More often, the economic beneficiaries are those who effectively use the new technology commercially.
Here again, diversification comes into play. Can we or anybody declare with certainty who those economic beneficiaries will be from a technology that can improve itself and likely someday will produce commerce we humans have yet to imagine? Broad
exposure to AI seems the correct posture. Broad enough to own the unidentifiable winners that will more than take care of losses from the companies left behind. Or, as Howard Marks puts it in his February 26 memo: “A moderate position, applied with
selectivity and prudence, seems like the best approach.”
We invite you to read his entire memo which we found illuminating and fascinating here:
Not a governing entity in the USA of remaining humans is truly addressing either of the two largest social questions before us. AI is one. The other? Government benefits in the form of Social Security and Medicare. According to the Social Security Trustees' projections, benefits could be reduced by approximately 15% starting around 2030 unless funding conditions change. (The key word in this sentence is “starting”, not the amount or the date). AI will likely help us with the math and various options. AI cannot, however, decide for us.
You also have diversification within your holdings away from the U.S. The broad index of global stocks we use as our benchmark for equities (MSCI All Capitalization World Index, “ACWI”) has performed better than the S&P 500® Index over the last year going back to early April. Some of that performance differential can be attributed to the tariff policy of the U.S. as well as U.S. dollar weakness against some other major currencies over that period.
- Emerging Markets stocks – the without-China version has done better than the with version
- Developed Markets stocks ex-U.S.
- Precious metals and oil
- Defense stocks
A broadening out in U.S. equity returns can be seen in U.S. MidCap 1
coming to life (+9.38% YTD) and the equally weighted version of the S&P 500® Index outperforming the benchmark cap-weighted version (+7.00% vs +0.67% YTD). That broadening out is healthy in our view, and helps pull some risk out of the stock market, at least in the U.S.
All of which is a very wonky way of trying to tell you that your investments are doing just fine when looked at as a whole. The total value of what you own in your investment account is rising, not falling as you may have thought if you watch financial news. While it is much too early to claim victory either over Iran or the attainment of our annual target returns, the news for now is favorable.
The Twilight Zone
You have travelled on new routes for the last 25 years, which has brought you to this place. A place intensely dark, seemingly impenetrable, perhaps forbidding.
You have travelled on new routes for the last 25 years, which has brought you to this place. A place intensely dark, seemingly impenetrable, perhaps forbidding.
This place may offer an abundance of wealth and possible solutions to the problems you brought with you.
In a forest, even a very dense one, there are many entry points. Here, are there paths? Are there hostiles or friendlies, or both?
Or do you bushwhack, and blaze your own trail to discover what can be uncovered in this new land?
Welcome to the shores of Artificial Intelligence.
Artificial Intelligence (“AI”) will likely transform our lives to at least an extent as did the Internet itself – upon which AI itself depends to be an actor. Its utility, in economic terms, will likely be a multiplier on the Internet itself.
Without electricity, AI does not exist. Without connectivity, AI does not exist. It is not as simple as unplugging that space heater that may burn your home down, but nor is it that different. So long as humankind controls both inputs, we should continue our existence. (Nod to HAL2000).
That is the “existential’ argument.
The internal argument is what do we do with Frankenstein, now that we have made him?
Dario Amodei, an OpenAi deserter who founded Anthropic with his sister and four others, worries about the jobs AI can do better, faster. The open question here is : cheaper? Sunk costs on at least $1 trillion in data centers has to be factored in. Yes, there is a reverse “J curve” possible: costs could plummet with wide adoption. But, as Davio worries, what is the social cost?
As commentators other than us have pointed out, wealth accrues to the economic beneficiaries of new technologies. Those beneficiaries are often not the inventor or the investors who invested in the inventor’s idea to bring it to life. More often, the economic beneficiaries are those who effectively use the new technology commercially.
Here again, diversification comes into play. Can we or anybody declare with certainty who those economic beneficiaries will be from a technology that can improve itself and likely someday will produce commerce we humans have yet to imagine? Broad
exposure to AI seems the correct posture. Broad enough to own the unidentifiable winners that will more than take care of losses from the companies left behind. Or, as Howard Marks puts it in his February 26 memo: “A moderate position, applied with
selectivity and prudence, seems like the best approach.”
We invite you to read his entire memo which we found illuminating and fascinating here:
Not a governing entity in the USA of remaining humans is truly addressing either of the two largest social questions before us. AI is one. The other? Government benefits in the form of Social Security and Medicare. According to the Social Security Trustees' projections, benefits could be reduced by approximately 15% starting around 2030 unless funding conditions change. (The key word in this sentence is “starting”, not the amount or the date). AI will likely help us with the math and various options. AI cannot, however, decide for us.
- Specifically, the Schwab Mid-Cap Index.
DISCLOSURES:
Past performance does not guarantee future results.
based on accounting procedures, reporting dates, or valuation methodologies of certain securities.
uniform.
in by you. Index performance is also for illustration only, to provide a relative comparison for the model portfolio. Index performance does not reflect any management fees, transaction costs or expenses.