How Do You Invest For A Crash And A Rally?
Inflation is back at 3.7%, and we think it is being pushed up by tariffs, war, a shortage of workers, big deficits and an AI spending race. And the AI build-out increasingly runs on borrowed money. That leaves investors with two stories running at once. Which one wins, and how do you invest when you can’t know?
In one story, the S&P 500 is up 12% this year and Nvidia sits about 9% below its Oct 6 record.
In the other story, Oracle is down about 60% from its September 2025 peak, yields on CoreWeave’s 2031 bonds have topped 15%, and since early October chip-equipment makers Lam Research, Applied Materials and KLA have each fallen about 20% to 24%, after losing streaks of nine or ten days.
So which story wins? The honest answer: a 50% fall is possible, but it isn’t our base case. Earnings down about 25% and the market’s price-to-earnings ratio falling from about 25 to about 16 would do it. Falls of 45% or more have happened three times since 1970: 1973–74, 2000–02 and 2007–09.
So here’s the better question. Not “which way will it break?” but “how do I invest so I’m fine either way?”
First, what breaks
When rates rise and money gets tight, the companies most likely to fail are the ones that need to borrow to survive. We scored 358 big US companies, each worth $5 billion or more with $5 billion or more of debt, on how well they can carry it: cash, cash burn, interest cover and debt load. Many look weaker than Oracle on paper, among them the new AI data-center builders, which went from almost no debt to billions in two years. Few are near failing.
And the pattern on the other side is just as clear. We compared five pairs of rivals, one cash-rich and one debt-heavy; three are below. In each pair, the likely winner has net cash or little debt, and the likely loser has a lot of debt:
| Industry | Cash-rich | Debt-heavy |
|---|---|---|
| Cloud and AI | Microsoft: $20B net cash; operating income covers interest 51×; stock 2% below its 52-week high | CoreWeave: $46B net debt, including leases; an operating loss, against $1.9B of interest; 67% below its June 2025 peak |
| Communications | Alphabet: $130B net cash, including $87B of stock it holds in other companies; operating income covers interest 66×; 14% below its 52-week high | Charter: $96B net debt; operating income covers interest 2.6×; down 57% in a year |
| Materials | Nucor: $4.4B net debt; operating income covers interest 29×; 5% below its 52-week high | Celanese: $11B net debt, about 2.1× its market value; operating income covers interest just 1.1×; 31% below its 52-week high |
Net cash and net debt are cash and short-term investments minus total debt, including leases. Interest cover is operating income divided by interest over the past twelve months.
And it isn’t just this year. From January 3 to October 12, 2022, the last time rates jumped, the stronger name did better, including dividends, in all four of the five pairs that were trading then. Nucor rose 7% while Celanese fell 46%. Alphabet fell 33%; Charter fell 53%.
Rule one is simple: in a squeeze, cash beats debt.
Why balance, and not a bet
Look at how stocks and the usual hedges, long bonds, gold and energy, did in three very different markets.
| What happened | S&P 500 | Long Treasuries | Gold | Energy stocks |
|---|---|---|---|---|
| 2022 (Jan 3–Oct 12): inflation and rate hikes | −25% | −29% | −7% | +45% |
| 2007–09 (Oct 9, 2007–Mar 9, 2009): a credit bust | −57% | +26% | +24% | −48% |
| 1995: a soft landing | +34% | rose (10-year yield 7.8% → 5.7%) | — | — |
See the problem? The asset that saved you in 2007–09, long bonds, sank you in 2022. The asset that saved you in 2022, energy, fell 48% in 2007–09. No single holding wins every storm. A mix that owns a little of each loses far less than the worst of them in any storm.
That’s what balance is: not a forecast, but insurance against being wrong.
Five ways to allocate differently now
Own quality, not just size. The S&P 500’s ten largest companies are 40% of its value. Owning the index now means a big bet on ten companies. Favor companies that could live without a bank: net cash, earnings that cover interest many times over. Trefis’s High Quality (HQ) strategy is built around selecting quality companies.
Get paid while you wait. A three-month Treasury pays 4.3%. Dividends from companies that keep raising them add income that doesn’t depend on guessing the market.
Spread your bonds across maturities. Long bonds lost 29% from January to October 2022 and gained 26% in 2007–09. Short ones moved far less: down 4% in 2022 and up 9% in 2007–09. Owning both means you lose less on either road: half long, half short lost about 17% in the 2022 sell-off instead of 29%, and still gained about 17% in 2007–09.
Don’t let one stock own you. If Nvidia or another winner has grown to a large share of your wealth, a 37% drop, which Nvidia has already had once in this boom, would take a big bite. Trimming on a schedule, and doing it tax-smart, is cheaper than trimming in a panic.
Rebalance by the calendar, not the headlines. Decide your mix, then reset it twice a year. It forces you to sell what’s run up and buy what’s fallen, which is hardest to do by feel.
Capping one large position without a big tax bill is the hardest of these five to do alone. See how a floor and a gradual exit work.
Stay calm: how storms end
Every storm on that list ended. From the March 2009 bottom, the S&P 500 rose 69% in a year. From the October 2022 low, it rose 22% in a year. The investors who came out fine weren’t the ones who called the top. They were the ones still invested, and still balanced, at the bottom.
Peace of mind doesn’t come from knowing what happens next. It comes from knowing your portfolio can take whatever does.
The verdict: don’t bet on the road; build for all of them.
Favor cash over debt. Get paid to wait. Spread bonds across maturities. Cap what any one stock can do to you. Rebalance on a schedule. Do that, and you can live through a 50% crash or another big rally.
The Hardest Move To Make On Your Own
Most of these five moves are simple to do yourself: favor quality, collect interest and dividends, spread bonds across maturities, and rebalance twice a year. One is not. Capping what one stock can do to you is hard when that stock has grown, over years of gains, into a big part of your wealth. Selling it all at once can bring a large tax bill, and holding all of it means riding out a crash with your plan tied to one company. There are ways in between: putting a floor under the position, paid for by giving up some of the upside, and moving out of it over time with less tax. Our partner advisory firm offers a free strategy review of how that would work for your holdings.