Interest Rates and Stocks: Right Call, Wrong Timing

An inverted yield curve can be right about direction and a year wrong about timing, and that gap is enough to lose money on a correct diagnosis. Which is why macro does not answer the question of what to do with a portfolio. It answers a different one: do interest rates, inflation or the currency change demand, costs, financing or the competitive position of this one company. If they do not, they stay background, and background is not a reason to move.

The signal that stayed on two years#

The classic warning went off on 6 July 2022. The spread between ten-year and two-year US Treasuries fell below zero that day, to minus 0.04 percentage points, in the series kept by the St. Louis Fed, and after a few attempts in late August it held above the line for good from 6 September 2024. Which means the light stayed on continuously for more than two years, rather than for the week after which you might reasonably do something.

The market, meanwhile, got on with it. The S&P 500 closed 2022 at 3,839.50 and 2023 at 4,769.83, according to the same institution's data, which is a gain of roughly a quarter in twelve months while the warning was still hanging over the market.

Anyone who treated the curve in January 2023 as an exit button gave that gain up for peace of mind. The recession did not arrive on that schedule, and the position was bought back higher, or not at all.

Why the curve works at all#

The curve inverts when the bond market prices future rate cuts, which means it expects an economy weak enough that the central bank will have to rescue it. Behind that reading sit the actual savings of thousands of participants rather than an opinion offered from a stage. That is where its weight comes from.

There is a more tangible mechanism too, and it shows up in an income statement. A bank borrows short and lends long, so it lives on the gap between the two ends of the curve. When the short end pays more than the long end, new lending stops being worth writing, credit growth slows, and companies that fund expansion with debt get less money on worse terms. This is not a correlation pulled off a chart. It is a cause sitting in the lender's income statement.

That is sound, and for most of post-war American history it held. Estrella and Mishkin measured the relationship across the post-war US recessions, and the curve owes its standing as a leading indicator to that work. Honesty requires the second half of the sentence: the lead there is measured in quarters rather than weeks, and the same warning has lit up when no recession followed at all.

Direction is not a date#

Except that direction and date are two separate pieces of information, and the market pays for the second one only.

Say the indicator is right and recession risk really is rising. What exactly do you do with that? Change the whole allocation, half of it, a fifth? Come back after a ten percent drop, after the first cut, after two quarters of better data? The entire decision sits inside those questions, and macro has nothing to say about any of them.

The warning says conditions are deteriorating. It never says in March.

Price the two-year mistake. Say you step out of the market and hold cash paying 4% a year, so after two years you have roughly 8% more than you started with. That is the whole return on the decision, which means the exit pays only if the market delivers less over the same window. It does not have to fall. It only has to grow more slowly, and those two years failed that test: 2023 on its own gained about a quarter. Then there is the second condition nobody prices at the exit: a day on which you recognise the bottom and go back.

On top of that, a share price moves between one set of results and the next for reasons that have little to do with the economy: fund flows, index rebalancing, somebody else's position being closed at the wrong moment. One thing is enough here: a macro headline and a price move can sit next to each other with no causal link at all.

A rate is the price of time#

Before macro reaches a company through demand, it reaches it through arithmetic, and that route is faster. A share price is the present value of future cash flows, so when the risk-free rate rises, the same future is worth less today. Nothing inside the business has to change for the number to change.

Say all the profit shows up ten years from now. At a 5% discount rate, a dollar from that year is worth about 61 cents today, and at 9% about 42 cents. Which means the same promise loses close to a third of its value in a single move by a central bank.

That is why rate rises hit companies whose profit is a promise harder than companies whose profit is a line in the accounts. The market is not judging their product or their management there, it is moving the reference point: a bond paying 5% starts competing with a share that pays hope.

Four questions instead of one#

Asked of the market, "does this change anything" has no good answer. Asked of one company, it has four.

• Demand. Does the customer buy this product on credit? A flat, a car and a production line, yes. A medicine, electricity and software paid out of an operating budget, far less often.
• Costs and margin. Does the company pass higher input costs into its price list the same quarter, or after a year of negotiating with three retail chains?
• Financing. When does the debt mature, and at what cost will it be rolled?
• Competitive position. Who in this industry cracks first when money gets expensive, and does your company pick up their customers or hand over its own?

The second question is settled in the price list, not in an inflation forecast. A company that raises prices in the same quarter its input costs rise comes out of inflation with its margin intact or better. A company that renegotiates once a year with three retail chains eats twelve months of cost increases out of its own margin, however warmly the chief executive talks about it on the call. The same inflation print is two different pieces of news for those two businesses.

The third question has the sharpest edge, because it can be counted. Say a company has $500 million of debt to roll next year. At a 3% cost the interest bill is $15 million a year; at 8% it is $40 million, so $25 million that used to stay inside the business now goes to the lender. No management presentation makes that go away.

The debt maturity schedule sits on the balance sheet, and it sits there long before the first headline about rates. You pull it out from the bottom, where cash and the debt due within a year sit.

The fourth question is the least intuitive, because harder conditions are not distributed evenly. When money gets expensive, the competitor with the weakest balance sheet cracks first: cutting investment, selling assets, losing customers. A downturn is therefore a test after which the leader can end up with a larger share of a smaller market, and no sentence about a rough year for the industry describes that.

A bank, a homebuilder, a fashion house#

The same rates land on three companies in three completely different places.

A bank lives on net interest margin, the gap between what it pays a depositor and what it charges a borrower. For JPMorgan, or for a large Polish lender like PKO BP, the level of rates is a line in the income statement, not a talking point.

A homebuilder takes the same move straight in the demand. Poland ran the sharpest version of that experiment: the monthly average central bank rate went from 0.23% in October 2021 to 6.68% in September 2022, in the central-bank rate series published by the St. Louis Fed, which means the cost of money rose nearly thirtyfold in eleven months. Mortgage instalments stopped fitting inside what buyers could borrow, and a homebuilder sees a move like that first in units contracted and only later in revenue, because revenue is booked when the flat is handed over.

A luxury house answers to a different judge. Before you decide that a headline about US rates touches a business like LVMH, open its annual report and check what share of sales goes to Asia. That one number tells you more about the next quarter than the statement after a central bank meeting.

A fourth company feels nothing at all. A maker of subscription software goes through the same year with almost no change in the business, though not necessarily in the valuation, because nobody borrows to pay a monthly invoice and nobody postpones one while they still need the program to work.

Which industry a company sits in decides how hard macro lands on it, and that part cannot be fixed with a better model, because the stage of the category is settled before the spreadsheet opens.

When macro really is the thesis#

The rule that macro is background has an exception, and it is better to name it yourself than to wait for the market to do it for you.

There are companies where macro is not background but the main variable. A bank, a homebuilder, a company with a lot of debt maturing soon, a commodity producer in the middle of a supply cycle. For those, the question of whether to leave the market is wrong for a different reason: the question belonged earlier, at purchase, and it was how much of the thesis rests on one interest rate.

Subtract that single macro factor from the thesis and look at what is left. If nothing is left except a hope for cuts, you own a bet on interest rates, priced as a share and with no settlement date.

The second risk is subtler. The phrase "it is only context" makes a convenient excuse for sleeping through a regime change. A company that grew mainly because money was free does not have worse sentiment after a hiking cycle. It has a different business model, and it has not put that in the deck yet.

Which is why, when a share price drops twenty percent, the first question is not what is happening to the market but which of my assumptions just stopped being true, and the answer fits into an hour written out as an order of operations.

Three scenarios instead of a forecast#

Since the date is unknowable, a macro forecast is the wrong tool and a scenario is the right one.

Write down three versions of the world: good conditions, average, hard. For each, answer what happens to demand, to margin, to the cost of debt and to customer behaviour. By the third version you can see how much of your thesis balances on a single assumption the company does not control, which is the whole point of writing them, because nobody picks the right scenario anyway.

In the hard scenario, do not guess the index. Answer something narrower: does the company close the year with positive operating cash flow, does it have to cut investment, does the covenant in its credit agreement survive a third off the result. Answers to those questions sit in documents rather than in forecasts.

I do not believe a thesis that works only if rates fall quickly, the consumer returns to spending, the currency helps and the competition leaves its price list alone. That is not a thesis. It is a wish list with four conditions, none of which the company controls.

Honestly: I do not know when the curve will invert next, or whether it will get the date right that time. I do know that I will put the question about timing to the company rather than to the curve.

The question from the other side, what moved the value of the portfolio over the last few days, is what Today's move answers in Taufolio, and sometimes the answer really is politics or macro data rather than any of your companies.

Write the condition, not the headline#

The practical difference between context and a compass is finally a difference in what gets written down.

The Investment thesis breaks the reason for holding a company into a few assumptions, each with a measurable condition. When the assumption reads "cost of debt below 5% after the 2027 refinancing", a headline about rates stops being an emotion and becomes a reading: the condition holds, or it broke. Monitoring then watches those conditions without you, on every plan and without spending credits, so the next headline about rates asks nothing of you until it moves one of the assumptions you wrote down.

What that looks like from the inside is on the product page.

Macro tells you what weather this company is sailing in. Whether the boat is watertight is something you have to read separately.