The Yield Curve — The macro signal that reaches directly into your business models

When Wells Fargo's net interest margin compressed from 5.0% to 4.6% between 2006 and 2007, the yield curve had already told you it was coming — the 10yr–2yr spread had inverted months earlier, quietly announcing that the bank's core profit engine was under structural attack.

That compression was not a surprise to anyone watching the right number. It was a predictable consequence of a business model — borrow short, lend long — running into an environment where that model stops working. The yield curve does not tell you what to buy. It tells you which questions to ask and which assumptions in your model are quietly wrong.

---

The concept in 60 seconds

A yield curve plots interest rates on government bonds from shortest maturity to longest. The number most investors watch is the spread between the 10-year Treasury yield and the 2-year Treasury yield — call it the 10yr–2yr spread, or T10Y2Y on FRED.

In a normal curve, long-term rates exceed short-term rates. This is the default, and most business models are built around it. A bank that pays 2% on deposits and earns 5% on a 30-year mortgage is living off that spread. The engine is called net interest margin, or NIM.

When the gap narrows toward zero, the curve is flat. When short-term rates exceed long-term rates, the curve is inverted. Inversion is historically rare. It has also preceded every US recession since 1978.

The practical upshot: the yield curve shape is not background macro noise. For any company that borrows money, lends money, relies on credit to grow, or is valued on distant future cash flows — the curve is a direct input to the business model.

---

Mental model

Think of the yield curve as a collective market bet on the future cost of money.

When the curve inverts, it is because investors are paying a premium to lock in long-term rates today. They are doing that because they expect short-term rates to fall — which happens when the economy slows and the Federal Reserve is forced to cut. The inversion is not evidence that something has already gone wrong. It is evidence that markets expect something to go wrong.

This timing distinction matters. The inversion is a leading indicator. It arrives months or years before the conditions it is pricing in become visible in earnings reports.

There is a second mechanism worth holding onto. When the Fed hikes aggressively — as it did from 2004 to 2006, and again from 2022 onward — the front end of the curve rises faster than the long end if markets believe the hiking cycle will ultimately damage growth. The curve does not wait for the damage. It prices it in ahead of time.

The practical mental model for analyzing a company: if the yield curve is inverted, your model's interest rate assumptions, credit availability assumptions, and discount rate assumptions all need to be stress-tested against a world where the damage is still coming, not already reflected.

---

Worked example: Silicon Valley Bank — March 2023

Silicon Valley Bank (SVB) is the cleanest modern demonstration of what the yield curve does to a bank that gets the duration math wrong.

In 2020 and 2021, the yield curve was steep. Short-term rates were near zero. Long-term rates were higher. SVB's deposit base surged as tech startups flush with venture capital parked cash with the bank. SVB deployed that cash into long-duration assets — US Treasuries and Agency mortgage-backed securities — to earn a spread. This was the classic borrow-short, lend-long trade, executed at scale, at exactly the wrong moment.

Starting in 2022, the Federal Reserve began one of the most aggressive rate-hiking cycles in modern history. Short-term rates rose sharply. The yield curve, which had been steep, flattened and then inverted. SVB faced the consequences from both ends simultaneously.

On the liability side: SVB's depositors — tech companies burning through cash — began withdrawing funds. To replace those deposits, SVB would have had to pay dramatically higher rates than it was earning on its long-duration asset book.

On the asset side: the market value of SVB's Treasury and MBS holdings collapsed. A bond bought at 1.5% yield is worth far less when the market is pricing equivalent bonds at 4.5%. SVB was sitting on billions in unrealized losses that the held-to-maturity accounting classification had kept off the income statement — until the bank needed to sell.

The yield curve inversion did not cause SVB's failure in isolation. The bank had a concentrated depositor base, poor interest rate hedging, and inadequate liquidity buffers. But the inversion was the mechanism that exposed every one of those weaknesses. The duration mismatch SVB had built in 2020–21 was survivable in a steep-curve environment. In an inverted one, it was fatal.

SVB was placed into FDIC receivership on March 10, 2023. In the preceding weeks, any analyst running a serious NIM sensitivity analysis against the prevailing curve environment would have seen the pressure building. The data was not hidden. The signal had been present for months.

---

Historical pattern

The track record is unusually clean by macro standards. Every US recession since 1978 — six recessions — has been preceded by a yield curve inversion. The 2yr–10yr spread inverted ahead of the recessions of 1980, 1981–82, 1990–91, 2001, 2007–09, and 2020.

The lag between inversion and recession onset has historically been 12 to 24 months. That gap is long enough that markets often continue rising significantly before the recession arrives — which creates two of the most common investor errors (covered below).

One pattern within the pattern deserves special attention: the re-steepening. After an inversion, the curve eventually begins to steepen again as the Fed starts cutting rates. This might look like an all-clear. Historically, it has been the opposite. The re-steepening — particularly a "bull steepener," where short-term rates fall faster than long-term rates — has often been a reliable early signal that the recession the inversion was warning about has arrived or is imminent.

The August 2007 re-steepening is the clearest example. The 2006–07 inversion had already flagged the credit stress building in the housing market. When the curve began to steepen in mid-2007, it was not because conditions had improved. It was because the Fed was beginning to respond to the deterioration. Investors who read that steepening as resolution were roughly 12 months ahead of the worst of the Global Financial Crisis.

---

Decision framework

This is a working tool for active research, not a macro forecasting exercise.

Step 1: Classify the current spread. Pull T10Y2Y from FRED. Above 150bps: broadly normal. Between 0 and 150bps: flattening — heightened attention warranted. Below 0: inverted — systematic review of NIM-exposed holdings required.

Step 2: Map your research candidates to the three exposure channels.

• Direct NIM exposure — banks, insurers, leasing businesses. For any company in this category, the curve environment is a direct earnings input. Run a NIM sensitivity analysis. If the spread compresses another 50bps, what happens to operating profit? The Wells Fargo 2006–07 example is the reference case: margin compression arrived months before the crisis.

• Credit-dependent growth — highly leveraged businesses, PE-backed companies, asset-heavy industrials financing expansion through debt. Ask directly: does this growth plan still make sense if refinancing costs rise 150–200bps? A business whose growth is funded by cheap short-term credit is operating a model that the curve environment can switch off.

• Long-duration valuation sensitivity — companies whose intrinsic value depends on cash flows projected 10–20 years out. Rising long-term rates raise the discount rate applied to those flows, compressing what a rational buyer should pay today. Rerun your valuation with the discount rate 100bps higher. If your conviction does not survive that test, the position sizing needs to reflect it.

Step 3: Flag the re-steepening. Add T10Y2Y to the event-driven monitoring list for any NIM-exposed position. A re-steepening after a prolonged inversion is not a signal to relax. It is a signal to revisit whether the thesis holds if the economic conditions the curve was pricing in are now arriving.

---

Common mistakes

Mistake 1: Acting immediately on inversion. The March 2019 inversion prompted many investors to reduce equity exposure in anticipation of an imminent recession. The S&P 500 ended 2019 up approximately 30%. Investors who sold on the inversion signal, or rotated heavily into defensive positions in mid-2019, missed most of that rally. The lag between inversion and recession is not a technicality — it is the dominant practical reality. The inversion tells you to stress-test your holdings and sharpen your awareness. It does not tell you to sell.

Mistake 2: Declaring the signal broken after a lag. By late 2019, the same commentators who had warned about the March inversion were being asked to explain why the recession had not materialized. Many concluded the yield curve had lost its predictive value — structural factors, global capital flows, central bank intervention had broken the signal. Then COVID arrived in Q1 2020 and the recession followed immediately. Whether the pandemic was the cause or the trigger for a recession already primed by underlying credit conditions is a legitimate debate. What is not debatable is that the curve had correctly identified the fragility. Calling the signal broken based on a 9-month lag was a category error.

Mistake 3: Reading the re-steepening as resolution. August 2007. The 2006–07 inversion had been well documented. When the curve began to steepen in the second half of 2007, some investors read it as the macro environment normalizing — the Fed cutting, the yield curve returning to a more favorable shape. What it actually signaled was that the Fed was responding to visible deterioration in credit markets. The recession officially began in December 2007. The worst equity drawdowns in the GFC came in 2008. Investors who treated the re-steepening as an all-clear and increased exposure to financials in late 2007 absorbed the full force of what followed.

---

How VI Stack uses this

The yield curve enters the research process at two points.

At Gate 3 (The Forensics), T10Y2Y is a direct input to discount rate assumptions in any DCF or terminal value calculation. An inverted or flattening curve warrants a sensitivity run at least 100bps above the base discount rate. For any bank or insurer under review, NIM sensitivity under the current curve environment is a required line item in the financial analysis.

In Block 4 (The Watch), T10Y2Y is a standing trigger on the event-driven monitoring list for every NIM-exposed holding. A re-steepening after a prolonged inversion is flagged as a thesis-review event, not a routine data point.

The goal is not to make macro calls. It is to ensure that macro reality is never invisible inside the business analysis.

---

What's next

The yield curve shapes the cost of money. What that means for how companies fund themselves — and how you should think about their cost of capital when building a valuation — is covered in the next module.

Macro 02 — Interest Rates and the Cost of Capital examines how the level of rates (not just the shape of the curve) flows through to WACC, hurdle rates, and the practical discount rate decisions you make inside a DCF.