the nangoy foundation
A personal record of one small book, run patiently.
september 2026 · book as of 2 october 2026
September continued the story August started, and the part that got worse was yields. The 10-year Treasury, which I left at around 4.72% at the end of August, crossed 5% in the middle of the month and closed September at 5.29%, its highest in a year — roughly half a percentage point in a single month. That came despite the Fed raising rates a quarter point on September 16th, to 3.75–4.00%, a unanimous decision justified by inflation still well above target and an oil price that won’t settle. Normally a hike calms the long end. This one didn’t.
The oil price is the Iran war, which is still with us. US and Iranian strikes resumed on the first weekend of the month, and Brent crude went from around $92 back above $100 before slipping toward $97 on talk of reopening the Strait of Hormuz. Nothing has been agreed. One analyst called it a “no war, no peace” situation that could run well into 2027, and that is how it feels.
The headline index hid what actually happened. The S&P 500 finished September down just 0.2% and the Nasdaq was up 2%, carried by semiconductors, which gained 11%. But the equal-weighted S&P 500 — where every company counts the same — fell about 5%. The average stock had a bad month; a handful of giants held the index up, and the ten largest now make up nearly 38% of it. I saw that split in my own book. Taiwan Semi rose almost 10% and Nvidia gained too, while Moody’s fell 10%, American Express 8% and Progressive 5%. By the start of October the tech side had already recovered; the rest of the book stayed stuck. That is what bad breadth looks like from the inside.
Then on Friday, October 2nd, the September jobs report came in weak: 29,000 jobs against roughly 84,000 expected, unemployment up to 4.2%, and the two prior months revised down by 60,000, leaving July with a net loss. Markets rallied anyway. The Nasdaq touched a record during the session, and the odds of another Fed hike in October fell to around 18%. Bad news for workers became good news for stocks, because it took the next hike off the table. Even that rally was narrow: the equal-weighted index rose less than half as much as the headline one.
I’ll be honest: I don’t know where the market goes from here. For me it’s a wait-and-see on two things — the Iran war, which runs straight through oil and into yields, and the November midterms. Until then I’m watching prices closely and looking for good opportunities, but I’m in no rush to put the cash to work.
Cash & T-bills combines the cash balance with the SGOV position — the true dry powder, a little over a quarter of the book. US equity now includes the IVV index holding. Emerging Asia is Tencent Music and Taiwan Semi; International developed is the VEA holding.
SGOV 0–3 Month Treasury · 13.4%Treasury bills — what I do with cash I’m not ready to spend. It earns a little while I wait, and I’m in no rush.
VEA Vanguard Developed Markets · 12.9%The un-clever part of the book on purpose: cheap, broad exposure to developed markets outside the US, and my bet that the whole world isn’t priced like the S&P.
MCO Moody’s · 10.2%A toll booth on global debt issuance, and one of the good businesses I like. It fell 10% in September after a strong run, and it will probably always trade at a premium for its quality. I’m happy with it as a business and don’t much care where the price goes in the short term — I didn’t sell on the way up, and I’m not panicking on the way down. I’d buy more on a bigger drop.
NVDA Nvidia · 9.3%A monster business at a price that demands near-flawless execution. I own it, I’m not adding, and I count it as part of one AI position rather than a separate bet.
AXP American Express · 8.7% · newA wide-moat compounder, bought at a price I set in advance. More below.
AMZN Amazon · 6.8%Really an AWS bet, and the quality work isn’t finished — I still can’t say whether AWS earns a good return on the capital it’s pouring in. Of the hyperscalers, it’s one of the two I like most.
IVV iShares Core S&P 500 · 6.7% · newThe S&P 500, in place of August’s QQQ. It puts some cash to work while nothing on my list is at my price.
TSM Taiwan Semiconductor · 5.7%The company the whole AI complex quietly runs on, and an elite compounder on returns. I own it, I’m not chasing it up here, and I’d want more on a broad selloff.
TME Tencent Music · 5.5%Priced for the business to shrink, which it isn’t. The risk that matters isn’t the numbers, it’s the China ownership structure behind it.
GOOGL Alphabet · 4.5%Still the honest oddity: my own valuation says pass. I haven’t settled whether the hyperscalers win long term, but of the group it’s one of the two I like most.
PGR Progressive · 2.9%A best-in-class insurer in a tougher patch. Its combined ratio — the share of premiums paid out in claims and costs — rose to 89.3 in August as prices per policy eased and claims cost more. That’s still well inside my 96% stop, and the price has moved closer to where I’d add.
Since inception on 3 July 2025, the book has returned +19.0% on a time-weighted basis — the clean measure of how the portfolio itself performed — against +24.0% for the S&P 500 over the same stretch. The gap widened this month: at the end of August the book was 1.8 points behind the index; it is now 5.0 points behind. That is the breadth split from the commentary showing up in the numbers — the index was carried by a handful of giants while the financials in my book fell.
Time-weighted return since inception (3 Jul 2025 – 2 Oct 2026).
since the last letter (september performance) — the breadth problem
Change from the last letter to this one (27 Aug – 2 Oct 2026); bars show the size of the fall. The headline index barely moved, the typical stock fell more than 5%, and the book landed in between. Equal weight measured by the RSP fund’s price.
for context — other markets, since inception
Most people file American Express next to Chase: a bank that issues credit cards, earns interest on balances, and lives or dies with the credit cycle. That’s the misunderstanding, and it’s why Amex is this month’s stock.
Amex doesn’t just issue the card. It runs the network the card travels on — what’s called a closed loop. When you pay with a Chase Sapphire, Chase is one link in a chain: the payment runs over Visa’s rails, and the fee the merchant pays follows Visa’s published schedule. Chase is a price-taker. When you pay with an Amex, Amex is the card issuer, the network and the merchant’s counterparty all at once. That gives it two things Chase can’t copy. The first is pricing power: Amex sets its own merchant fee, around 2.3%, directly. The second is data from both sides of the transaction — it knows who the cardmember is and where they spend, which is exactly what it sells back to merchants as access to an affluent customer base.
Chase is actually the bigger card issuer by purchase volume. It doesn’t matter. Being bigger on someone else’s network isn’t the same as owning one.
The lending is real, but it’s the sidecar — the cyclical, commodity part of the business. Most of the economics come from spending and fees, and the returns show it. Return on equity has run in the low-to-mid 30s every year except 2020, and each dollar Amex keeps has earned about 34 cents of new profit, while it still hands back around three-quarters of what it earns. Card fees are up 16% this year, and Gen Z now makes up 30% of new accounts worldwide. The funnel is refilling with the next generation of cardmembers.
I did the work on Amex in July and concluded it was a wonderful business at roughly fair value, so I didn’t buy. Instead I decided in advance the prices at which I would, each carrying a margin of safety. In September, as financials sold off on rising yields, the first of those prices was hit. I didn’t have to decide anything in the moment; I’d decided in July.
That isn’t the whole story, and I won’t pretend it is. In mid-September, the same day the Fed hiked, Amex’s CFO said spending by its small-business customers won’t turn around this year — and that’s exactly the segment Amex’s new small-business banking push is aimed at. I don’t have an answer to that yet. None of the three things that would make me sell has moved: return on equity sliding below 25% outside a recession, pricing power eroding, or younger customers no longer joining. The next real read is third-quarter earnings on October 23rd, and we’ll see then.
In the meantime, I’m happy to own it for the long term, and content with where the business is.
September was a clean-up month: three positions left the book and two came in.
Walmart I was happy to hold, but I’d never done the real work behind it, and a position without work behind it starts to mess with your psychology. I’d set myself a deadline to do that work or sell; the deadline passed, so I sold — later than I should have, but before it got into my head.
McDonald’s went into the too-hard pile. The brand may well be fine, but at around $252 it sat well above my own estimate of value, and the questions that would settle whether its traffic problem is temporary can only be answered with numbers McDonald’s doesn’t publish. Expensive and unknowable is a combination I walk away from.
Gold was sunk cost. I bought it a long time ago as a hedge against my equity exposure, when gold was at its highs, and without a real thesis behind it. Owning something because I already own it isn’t a reason, so it went.
On the other side, American Express came in at a price I’d decided on in advance — more on that in the stock of the month. In effect, the money from those three exits went into Amex, and I’m happy with that trade. I also bought IVV, the S&P 500 index, in place of the QQQ I sold in August. With this much cash and nothing on my list at a price I like, I’d rather put some of it to work in the index than force a single stock. Cash and T-bills are still about 27% of the book.
One thing I didn’t change: Alphabet stays, even though my own valuation says pass. I’m still working out whether the hyperscalers are good long-term prospects at all. Of the group, Alphabet and Amazon are the two I like most.