The Macro Investing Tool

Let’s start with how the move breaks down...

Since the start of September last year, the 10-year is up 89 basis points, and the real yield is up 96, i.e., that's the whole move and more. Breakeven inflation, the market's own forecast of inflation, has actually fallen from 2.41% to 2.34%. If this were a loss of confidence in American debt, the fiscal dominance nightmare, breakevens, would be rising, and the dollar would be falling, and neither is happening. The market believes the Fed and is demanding a higher real return because growth is strong and supply is heavy.

So what growth is the bond market pricing? The real 10-year yield is 2.83%, and the real rate five years forward, the market's best guess at the long-run neutral rate, is 3.0%. Real yields tend to sit around trend real growth over time, so the bond market is now pricing an economy that grows at roughly 3% in real terms, against an average of 1.5% a year from 2007 to 2019. That's a doubling of trend growth, and it's not a fantasy, because productivity is already up 2.2% on the year and has run at about 2.1% a year since 2019, and the Atlanta Fed's tracker has the current quarter growing at 5% in real terms.

So put fair value together: if inflation settles around 2.3%, which is where breakevens sit, and real growth runs at 3%, then trend nominal growth is a little over 5%, and that's roughly where a 10-year yield should sit. At 5.2%, the 10-year isn't high; it's fair value for the economy the market thinks is arriving. Nominal GDP is growing 6.6%, so the rate on the debt is still below the growth rate, which is what stops the debt burden from running away from you, and that's financial repression doing its job.

Equities are pricing exactly the same thing...

The Nasdaq is up 20% on the year and at a new high, and the reason is the same growth. Consensus has S&P 500 earnings growing about 15% next year on revenue growth of 11%, i.e., the equity market is pricing a growth acceleration too, and so is the bond market. A valuation is cash flow divided by the discount rate minus growth, so if rates rise because growth is rising, the gap doesn't move, and neither does the multiple, which is why the stock market hasn't cared.

I've been asking myself whether it's inflation rather than rates, which is the larger discounting mechanism for long-duration stocks, and I think it is. The 1970s multiple crush happened with negative real rates, because inflation did the damage, and the late 1990s were the mirror image, with real yields of 3.5% to 4% and expanding multiples because inflation was low and stable. So high rates without inflation being particularly strong isn't really an issue for cash-generating businesses, and the line that would change my mind is core inflation above 3%.

And inflation is going the right way. Core CPI is 2.45% on the year and 2.0% annualised over three months, so the momentum is running below the annual rate, which is what disinflation looks like while it's happening.

Core ex-shelter, what we call supercore, strips out the rent component that lags everything else by a year or more, so it's the cleaner read of underlying inflation, and it's sitting at 2.0%, i.e. right on target.

Productivity is up 2.2% while unit labor costs are up only 1.4%, which is the Greenspan tell, i.e., productivity absorbing wage growth so the economy can run hot without the inflation. That's why the analogy I keep coming back to is 1997, not 1999: a Fed that hikes once into a productivity boom and then stops, not one that keeps hiking until it breaks the cycle. Warsh hiking is still uncomfortable, but a single hike into an energy shock with contained unit labor costs is 1997.

The supply side is real, and it's where the term premium comes from. The extra yield investors demand for holding duration has roughly doubled to just under 1%, because private compute is now issuing debt on a sovereign scale to fund the AI buildout, and that's what it costs. Credit is pricing the weaker borrowers rather than the system, i.e., high-yield spreads are about 15 basis points wider on the month while investment-grade spreads are a touch tighter, and a race that neither the US nor China can afford to lose is one the state ends up underwriting.

Which brings me back to the front end and the curve...

The measure of financial conditions to follow is still the one built on 3-month rates and the dollar rather than the 10-year, because markets are following the short end, and the Nasdaq versus that measure is the chart that called the end of the sideways range two weeks ago. The caveat is that it has tightened since the Fed hiked and the dollar went back through 100, which was the condition I set two weeks ago, so at the next monthly read, the dollar probably joins the long end and oil as a third leg squeezing. In a hiking regime that resolves through the curve, the curve only steepens if the front end comes down.

The flatness in the curve is the 2-year. On Friday, the 3-month bill was 4.24%, the 2-year 4.81%, and the 10-year 5.17%, so 2s10s was only 36 basis points while 3-month to 10-year was 93. The 2-year is kinked because it's carrying the oil-driven hike pricing, and if we strip that out, the curve is already steep. I think the 2-year comes back down, the curve steepens, and that ignites the next leg of the funding cycle, because a steeper curve is what makes the banks buy Treasuries and lend against them. Put simply, the banks have about a trillion dollars of room, and they haven't used it yet, and that's the liquidity that funds the whole buildout without the Fed having to use its balance sheet.

The curve is already trying to turn. 2s10s bottomed at 20 basis points on the 21st and was back at 36 by Friday, a violent bounce off the floor of the channel it's been in since January, and the level to watch is 45 basis points, where the top of that channel meets the 200-day average. The dollar is the other lever, because if Bessent can get it lower, through liquidity, the swap lines, and intervention, that's the hypercharge for everybody's growth, and the Xi visit extended the trade truce with no big breakthrough, but I think the warmth matters.

Crypto has run ahead of this. Bitcoin is at about $84,500, but crypto is funded out of excess liquidity — the liquidity left over once the real economy has had what it needs, and our Excess Liquidity Index is still below zero and got worse in September. The ignition we're waiting for is two positive monthly prints in a row, which hasn't happened yet, and will require the same steeper curve.

The year-on-year version of the same picture, excess narrow liquidity against bitcoin's annual change, shows how tight that relationship is, and it says bitcoin's rally has run ahead of the liquidity, so there's nothing to chase.

Where am I wrong?

My own line for the state losing control of the long end was a real 10-year yield above 2.75%, and it has closed above that since the 23rd. Three closes don't make it, and with breakevens falling, the term premium under 1% and credit not stressed, this looks like a growth repricing rather than a buyers' strike, but if it holds there while the term premium climbs towards 1.5% and the dollar weakens with it, this whole read is wrong. The other risk is the Fed's own dots, because two more hikes with oil staying high will turn 1997 into 1999, with the 2-year going above 5% while the curve flattens before it steepens.

So, to sum it up, yields are at fair value for an economy growing 3% in real terms, not high, and both the bond market and the equity market are pricing that growth. That's no threat to the debt as long as nominal growth runs above rates, and no threat to equities as long as core inflation stays below 3%. The way through is a Fed that stops, a curve that steepens and banks that start lending against it, and that's the liquidity that funds the whole race.

Into next week, I'm watching the ISM on Thursday and whether new orders hold up, the 2s10s curve against 45 basis points, the dollar and whether it gets back under 100, the Iran road map and oil, and the real 10-year yield around 2.75%.

Raoul