🌾 Welcome to StableBread's Newsletter!
There’s a lot of market uncertainty right now, between the Iran war, oil prices swinging, and growing doubts about how long big tech can keep borrowing to spend on AI.
And if you’ve invested heavily in AI infra, rare earths, and/or defense, your portfolio has probably seen much better days.
Here’s my stock portfolio YTD, from just under 200% to now 64.8%:
Fajasy’s Stock Portfolio YTD Performance
Many investors, based on my observations and the emails I’ve received, have a really poor or surface-level understanding of what’s happening in the market.
They don’t know whether to buy, sell, or hold more cash, and that leads to irrational behavior and poor investment decisions.
So instead of writing about another obscure stock idea, I’ve decided to write about how to actually read the macro environment. I’ll also share which of my 30+ past write-ups offer the best risk/reward setups, and which I’ve already bought at discounted prices.
Let’s start with a short recap of the main headlines:
The S&P 500 closed at 7,580 on October 30, just 3.1% below its October 6 record.
WTI crude rose 15.2% in October to ~$104/barrel, as reports of U.S. plans to strike Iran after the midterms added to attacks on tankers near the Strait of Hormuz. It then fell 7.4% on November 2, after the U.S. and Iran signaled they’re open to talks.
The 10-year Treasury yield hit 5.31% on October 5, its highest close since May 2002, before easing to 4.94%.
And the Fed held rates at 3.75-4.00% on October 28.
So how do you make sense of all of this? Here’s where I looked for answers:
Growth × Policy: A map I use that indicates whether the Fed’s rate is high enough to slow the economy. It pairs a growth score (the ISM Services PMI and industrial production) with the real fed funds rate, which is the Fed’s policy rate minus core PCE inflation.
Gave’s Four Quadrants: Charles Gave’s framework (from Gavekal), which uses two price ratios, S&P 500/WTI crude oil and gold/IEF (a 7-10 year Treasury ETF), to place the economy in a boom or a bust phase, and tell whether it’s inflationary or deflationary. Gave developed it in 1978, after watching Nixon break the gold standard and markets descend into chaos, and has refined it through five decades of market cycles.
Credit × Duration: A map I use that splits bond moves into interest rate risk (the 10-year Treasury yield) and default risk (the high-yield spread, the extra yield junk-rated companies pay over Treasuries). Rising Treasury yields mostly hurt long-term bonds and growth stocks, while a wider spread hurts junk bonds and the companies that need to borrow.
CAPE Ratio: Robert Shiller’s measure of how expensive U.S. stocks are. It divides the S&P 500’s price by its average inflation-adjusted earnings over the last 10 years. I also look at the Fed model and Shiller’s excess CAPE yield, which corrects the Fed model for inflation.
Bubble Triangle: William Quinn and John Turner’s framework, which says a bubble needs three sides: marketability (how easily an asset trades), money and credit, and speculation. When a bubble bursts, the damage depends on how tied the bubble is to the rest of the economy, most often through banks. It’s especially important to check given today’s AI infra spending and the “AI is a bubble” claims.
Comparing where these frameworks agree and disagree is more informative than evaluating them on their own. It also gives you a much clearer read on how to position your portfolio.
P.S. The readings discussed in this post can be seen on my free macro dashboard, and I explain each framework in more detail on my YouTube channel.
Growth Is Still Positive
Growth × Policy reads growth (whether the U.S. economy is expanding or shrinking) from the economic data, while Gave reads it from market prices.
Both still point to growth:
Growth × Policy: The ISM Services PMI, a survey of the managers who buy supplies for U.S. services companies, came in at 54.9 in September (above 50 means expansion), and October’s reading comes out November 4. Industrial production rose 2.3% y/y in September, up from 1.6% in August and above 2% for the first time since September 2022. So factory, mine and utility output is now adding to the growth score instead of dragging it down.
Gave’s Four Quadrants: The S&P 500 was worth 72.8 barrels of WTI crude at the end of October vs. its 7-year average of 71.6, down from 85 at the end of September as oil jumped and the S&P 500 fell 0.9%. When stocks outrun oil, Gave reads that as a boom, since companies are getting more value out of the energy they use.
Real GDP also grew at a 3.4% annual rate in Q3, up from 2.2% in Q2, and Growth × Policy still places the U.S. in its risk-on phase, with growth above trend and the real rate below 1%:
Growth × Policy (StableBread Macro Dashboard; October 30, 2026)
While economic data still points to growth, oil left Gave’s market-price reading just 1.7% from the line between an inflationary boom and an inflationary bust (see visual below). In other words, a ~1.7% drop in the S&P 500 or rise in oil would’ve flipped Gave’s reading to a bust at the end of October.
That’s how fast the reading swings with oil. When oil fell 7.4% on November 2, the ratio jumped to 78.2, ~9% above its 7-year average.
Inflation Is Sticky
Inflation shows up in both the government’s price indexes and the price of gold:
Growth × Policy: Core PCE inflation, the PCE price index minus food and energy, rose to 3.1% in September from 3.0% in August, 1.1% above the Fed’s 2% target. CPI inflation rose to 3.7% from 3.4%. I focus on core PCE because the Fed sets its 2% target on the PCE index, and leaving out food and energy, whose prices swing the most, provides a better read on where inflation is headed.
Gave’s Four Quadrants: At the end of October, an ounce of gold bought 42.5 shares of IEF vs. a 7-year average of 26.3. That’s down from 47 at the end of September, since gold fell 6.9% in October while IEF rose 3.1%. Gave reads gold beating bonds as the currency losing value, and gold has been beating bonds since March 2020.
Gave’s Four Quadrants (StableBread Macro Dashboard; October 30, 2026)
Gave’s framework still places the U.S. in an inflationary boom, where the U.S. has been since November 2023, apart from four weeks in March and April 2026.
That’s the quadrant where he favors gold, commodities, real estate, value stocks, emerging markets and cyclicals with high fixed costs, and avoids long-duration bonds. These assets either rise in price with inflation or earn more when the economy runs hot, while long-duration bonds pay a fixed amount that inflation eats away.
High yields don’t change that call, since Gave’s gold/IEF ratio divides the gold price by IEF’s total return (interest included), and the ratio is still 61% above its 7-year average even after October.
Rates Eased, Spreads Widened
The 10-year Treasury yield, which mortgage rates and corporate bond yields are priced off, eased after its October 5 peak but is still near its highest level since 2002:
10-Year Treasury Yield (FRED; U.S. Treasury)
The media mostly talks about the 10-year yield and the Fed, but in October, the bigger shift was in credit:
Credit × Duration: Over the 13 weeks to October 29, the 10-year still rose 0.23% despite easing from its October 5 peak, and the high-yield spread widened 0.41% to 3.26%. All 7 weekly readings since September 18 have been bond selloffs, with yields and spreads rising together. That means borrowing got more expensive for everyone, and the most for junk-rated companies, which pay both the higher Treasury yield and the wider spread.
Growth × Policy: The Fed held its target at 3.75-4.00% on October 28, with 2/12 members voting for another 0.25% hike. With core PCE at 3.1%, the real fed funds rate slipped to ~0.8% from ~0.9%.
Credit × Duration (StableBread Macro Dashboard; October 30, 2026)
Every U.S. recession since 1970 except COVID began while the real rate was above 1%, and the current ~0.8% is just below that line.
Granted, a real rate above 1% doesn’t guarantee a recession. Since 1980, the real rate has been above 1% a little more than half the time, including every month from 1995 to 1999.
U.S. Stocks Are Expensive
By the CAPE ratio, U.S. stocks are close to the most expensive they’ve been since 1881:
S&P 500 CAPE Ratio (Robert Shiller; BLS; Yahoo Finance)
Valuations look stretched on their own and against bonds:
CAPE ratio: 40.6x in October 2026, more than double its 18x average since 1881. The CAPE has only been higher in 1999-2000 and August 2026 (41.0x). Since 1881, a CAPE of 30x or above was followed by -1.0%/year after inflation over the next 10 years on average, and investors lost money 3/5 times. Below 10x, the S&P 500 returned 11.3%/year after inflation, and nobody who bought at those levels lost money over the next 10 years.
Fed model: Stocks’ earnings yield vs. the 10-year Treasury yield. The S&P 500 traded at 18.7x forward earnings at the end of October, a 5.35% earnings yield, 0.41% above the 10-year’s 4.94%. The two were about even at the end of September. On the CAPE’s 2.46% earnings yield (1 / 40.6), the 10-year’s 5.13% October average wins by 2.67%, the widest gap since November 2000. So on next year’s expected earnings, stocks barely pay more than bonds, and on their 10-year average earnings, they pay far less.
Excess CAPE yield: 0.71% in October, its lowest since March 2002. Unlike the Fed model, it compares stocks and bonds after inflation, since stock earnings tend to rise with inflation while a bond’s interest payments don’t. Shiller measures the bond side as the 10-year yield minus the past 10 years’ average inflation, not the TIPS yield, which is why it’s still positive. Since 1881, readings below 2% were followed by 4.1%/year after inflation over the next 10 years on average, vs. 10.4%/year at 6% or more.
Concentration: Info tech was 39.7% of the S&P 500 at the end of September, the latest breakdown S&P has published, above its ~33% peak in 2000. That makes an S&P 500 index fund a ~40% bet on tech at a CAPE above 40x.
Not to mention, the 10-year yield on TIPS, the Treasuries whose principal rises with inflation, averaged 2.78% above inflation in October, more than the S&P 500’s 2.46% CAPE earnings yield for a second straight month.
September was the first time that happened since at least 2003:
10-Year TIPS Yield vs. the CAPE Earnings Yield (FRED; U.S. Treasury; Robert Shiller)
In other words, a 10-year TIPS bought in October locks in ~2.8%/year above inflation if held to maturity, with no default risk, which is more than the S&P 500 earns on its 10-year average profits.
Although a high CAPE indicates U.S. stocks are expensive, it doesn’t mean stocks are about to crash. The ratio was above 30x in 80 of the 120 months after September 2016, and the S&P 500 still returned 11.6%/year after inflation over those 10 years.
Instead, it implies lower expected returns from here.
About 4.2%/year of that 11.6% came from investors paying more for each dollar of earnings as the CAPE rose from 26.7x to 40.4x in September 2026. So repeating that over the next 10 years would take a CAPE of ~61x, clearly far above the 44.2x record from December 1999.
AI Boom Runs on More Debt
The companies spending the most on AI are paying for more of it with debt:
Debt Increase as a Share of Capital Spending (FactSet, July 23, 2026)
To understand whether debt-funded spending like this is indicative of a bubble, we can look at Quinn and Turner’s bubble triangle:
The Bubble Triangle (StableBread)
The AI boom arguably has all three sides:
Marketability: Index funds held 52% of U.S. long-term fund assets at the end of 2025, up from 19% in 2010, and they buy each stock in proportion to its size, so the biggest AI stocks get the most of every new dollar.
Money and credit: Microsoft, Alphabet, Amazon and Meta are on track to spend more than $730B on capex in 2026, up from $410B in 2025, and their Q3 reports didn’t change that. Amazon still expects ~$220B, and Meta narrowed its range to $135-145B. Investors are borrowing more too, with margin debt at $1.48T in September, just below June’s record $1.50T.
Speculation: AI startups raised 82.7% of the money U.S. venture capital firms invested in 2026 through September, a record share, up from 65.4% in 2025. So more of the private money is betting on a single theme.
Meta’s also raising money by selling stock. Its October 27 pricing term sheet shows it sold 40M shares at $625 through forward sale agreements, plus $12.5B of mandatory convertible preferred stock, which together bring in $37.4B of net proceeds once the forward sales settle by November 2027.
Worth highlighting that as a share of the U.S. economy, margin debt is now well above its dot-com peak:
Margin Debt as a Share of U.S. GDP (FINRA; FRED)
Quinn and Turner found that tech bubbles with little borrowing, like the dot-com bubble, did little harm to the economy. The ones that ran on borrowing, like the 1840s railway mania and the 1920s, ended in financial crises.
With a growing share of AI spending funded by bonds and bank loans, the AI boom is moving toward the ones that ran on borrowing:
Bubbles by Spark and Borrowing (William Quinn and John D. Turner, Boom and Bust)
That doesn’t mean a bust is coming. But if one does, the losses won’t stop with stock investors like they mostly did in 2000, since more of the spending is now owed to bondholders and banks.
Note: U.S. household debt was ~66% of GDP in mid-2026, its lowest since 1998, so the borrowing this time is in corporate debt and margin accounts, not mortgages and consumer loans.
Mixed Signals on Cash and Bonds
The readings we’ve discussed don’t all agree on what to do with cash, high-yield bonds and long-term Treasuries:
Cash: Growth × Policy’s risk-on phase favors stocks and real estate over cash, and all 13 of its weekly readings since August are risk-on. But in 2022, when Gave’s framework placed the U.S. in an inflationary bust, cash was one of the few assets that didn’t lose money.
High-yield bonds: Risk-on also favors high-yield bonds. But at 3.26%, the bonds pay less over Treasuries than their 3.60% median since 2019, so investors are being paid less than usual for the risk of default, even after October’s widening.
Long-term Treasuries: Gave avoids long-duration bonds in an inflationary boom. However, if growth slows while the real rate stays under 1%, Growth × Policy moves to a soft landing, which favors long-term Treasuries.
With T-bills paying ~1% more than core PCE inflation and junk bonds paying less than usual for their risk, I’d rather hold cash than high-yield bonds. And I’d wait for Growth × Policy’s growth score to turn negative before buying long-term Treasuries.
What Would Change the Readings
Two paths would change these readings the most:
2022 path: Oil spikes, and the S&P 500/WTI ratio drops below its 7-year average and stays there, which Gave calls an inflationary bust. The last long inflationary bust was 2022, when the S&P 500 lost 18% and the main U.S. bond index lost 13%, while gold was flat and XLE, the S&P 500 energy ETF, returned 64%.
Credit path: High-yield bonds’ extra yield over Treasuries rises while Treasury yields fall, which Credit × Duration calls a flight to quality. If more of the AI debt has landed on banks by then, the damage would look more like 2008 than 2000.
October moved the U.S. closer to both. The S&P 500/WTI ratio fell from 85.1 on October 9 to 72.8 on October 30, its smallest cushion over the 7-year average since it dipped below for four weeks in March and April 2026.
And from October 5 to October 22, the 10-year yield fell from 5.31% to 4.96% while the high-yield spread widened from 3.12% to 3.40%. That’s what a flight to quality looks like, but it lasted less than three weeks, too short to change Credit × Duration’s 13-week reading.
Gave’s early warning for the 2022 path has already flashed. He says stocks tend to fall behind gold a few months before they fall behind oil, and the S&P 500/gold ratio has been below its 7-year average since the end of January 2025.
From then to the end of October 2026, the S&P 500 rose 25.5%, but gold rose 38.6%. So measured in gold, the S&P 500 fell 9.5% (from 2.15 oz. of gold to 1.94), and the drop was 15% at the end of September, before gold slid in October.
If stocks also fall behind oil, they’d be losing ground to both, which is Gave’s inflationary bust.
How I’d Position
Here’s what I’d do with all of these readings:
Less in U.S. index funds: If the CAPE’s history holds, U.S. stocks will return less over the next 10 years than the S&P 500’s 7.1%/year average after inflation since 1871. That doesn’t mean you should stop dollar-cost averaging into your retirement accounts or sell what you own, but there are better risk/reward opportunities than an index this concentrated in U.S. large-cap tech.
Value stocks, emerging markets and real assets: Gave favors all three in an inflationary boom. Growth × Policy favors emerging markets and real estate in its current phase, and value stocks, emerging markets and commodities if the real rate rises above 1% (it’s ~0.8% now, and headed above 1% if the Fed hikes again in December). According to Barclays, whose CAPE for U.S. stocks is 39x, the UK (18x) and China (14x) trade at less than half that.
TIPS over nominal Treasuries: In October, 10-year TIPS yielded 2.78% above inflation on average and nominal 10-year Treasuries yielded 5.13%, so TIPS come out ahead over the next 10 years if CPI inflation averages more than 2.35%/year (5.13% minus 2.78%). CPI inflation was 3.7% over the 12 months to September, so inflation would have to slow well below today’s pace for nominal Treasuries to win.
Gold over nominal Treasuries: Gave favors gold over bonds for as long as the gold/IEF ratio stays above its 7-year average. It’s 61% above it now, so gold would have to lose ~38% against IEF to flip that reading.
More cash than risk-on alone suggests: Risk-on underweights cash, but with the S&P 500/WTI ratio swinging with every Iran headline, I’d keep more, since Gave’s inflationary bust favors cash and energy and avoids long-duration stocks and bonds.
In short, I’d own less U.S. large-cap tech, more of what an inflationary boom favors, and enough cash to act if Gave’s reading flips to a bust.
What I’m Watching
These would change the read the most:
Fed’s December 8-9 meeting and core PCE: Another 0.25% hike with core PCE at 3.1%, or core PCE below 2.9% with the fed funds rate unchanged, pushes the real rate above 1%, which Growth × Policy calls late cycle. That favors commodities, value stocks, emerging markets and banks over long-term Treasuries and growth stocks. Futures put the odds of a December hike at ~75%, and October’s core PCE comes out November 25.
S&P 500/WTI vs. its 7-year average: If the ratio falls below its average and stays there, the U.S. is on the 2022 path. That means Gave’s framework would switch from favoring gold, commodities and value stocks to favoring cash and energy.
High-yield spread: 3.26% on October 29. If the spread keeps rising while the 10-year yield falls for longer than three weeks, the U.S. is on the credit path. It’s a sign investors expect more defaults, and during the last flight to quality in early 2020, long-term Treasuries returned 22% while high-yield bonds lost ~12%.
Bank-held AI debt: In September, the Financial Times reported that ~$18B of bank loans for Project Jupiter, a data center Oracle is leasing in New Mexico to serve OpenAI, were being quoted at 89-91 cents on the dollar, and the banks were stuck holding more of the debt than they had planned. At its October 28 Investor Day, Oracle said its capex peaks in FY2027-2028 and steps down in FY2029 (years ending in May), when it expects at least $30B of free cash flow.
Right now, the readings point to growth with sticky inflation, and both paths that would end it are closer than they were at the start of October.
What I’d Buy Now
Here’s which of the stocks I’ve written up I’d buy, based on the readings above and each stock’s November 2 close.
Subscribe to StableBread's Research!
Full access to this write-up and every future one. Cancel anytime.
View PlansSubscribe to unlock:
- 50+ write-ups every year
- Buy/sell/hold decisions
- Periodic industry, market, and thematic research
- Deeper analysis and valuation models
- Full archive of all past research
