*Disclaimer: I just bought BRK.B so my analysis may be biased. Data sourced from Tradingview, Xfinlink and FRED. This article is not meant to be financial advice, DYOR.*
# tldr;
US debt is too big for the old "print and normalize later" playbook, so the next crisis likely gets solved the British way: rates below inflation for years, bondholders eat the loss. In that world real assets win. BRK is the cleanest way to play it: $300B cash hedge, insurance float leverage, railroad/utilities/energy, plus careful AI exposure through Alphabet/Apple and the power grid. Market prices it for \~4% growth vs \~9% for the average SPY stock, so the bar is on the floor. Wins if AI booms, buys the wreckage if it pops.
# Why BRK.B?
Since 2008 every crisis, big or small, has been solved the same way: Fed prints, markets recover, everyone moves on. But that playbook only worked because of two conditions nobody noticed at the time. Government debt was low, and inflation was dead. Printing was free.
Neither condition exists anymore. Debt is past 120% of GDP and spending keeps climbing under both parties. At that size, the government can't keep rates above inflation for long. The interest bill alone would eat the budget. We actually got a live demo of this: the Fed took rates above inflation in 2022-23, and the result is an interest bill that's now blown past $1T a year and rising. And since a chunk of today's inflation is supply-driven (oil, shipping), killing it would take Volcker-sized hikes held through a recession, exactly the kind of bill the budget won't be able to pay. Which means the one tool that kills inflation is the one tool they can't afford to use.
The UK hit this exact wall after WW2, and their solution is probably our preview. They printed anyway, pinned rates below inflation for \~30 years, and let inflation quietly shrink the debt. It worked. The cost was carried entirely by savers and bondholders.
If that's the game (and I think it is), the Fed's next move barely matters. Hold, hike, panic cut, whatever. Every path ends with inflation running hotter than rates. And in that world, real assets and pricing power beat paper claims on fixed dollars. Bonds are the designated loser.
Then there's AI, which complicates things. Biggest capex boom of our lifetime. If it delivers you don't want zero exposure. If it's a bubble, the unwind IS the crisis that kicks off the printing I just described. So ideally you want something that wins if AI works AND has dry powder if it blows up. Most portfolios force you to pick one.
So what to buy? You could just buy hard assets but gold and other commodities have already run hard, and they give you limited AI upside.
Which brings me to Berkshire. There are four points about BRK that makes it a solid pick.
1. The cash pile is the hedge. \~$300B sitting in short term treasuries collecting yield. No puts, no fancy hedging, nothing. Yes, in my own scenario that cash bleeds a little purchasing power every year. That's the carry cost of the option, and I think the option is cheap. When something breaks, Berkshire is the one writing checks at fire sale prices. They did it in 08, they'll do it again.
2. Insurance float is basically permanent leverage at near zero cost, and rising rates fatten the investment income on it. The flipside is claims inflation, which hits costs too. BRK's underwriting discipline has handled that better than almost anyone (GEICO took the hit in 21-22 and fixed it), but it's income with a caveat, not a free lunch.
3. The operating businesses are real assets. A railroad, utilities, energy. The railroad reprices with inflation almost immediately through fuel surcharges and contract resets. The utilities reprice slower since regulators have to approve rate hikes, but they get there, and meanwhile they own the one thing AI needs no matter what: electricity.
4. It's the careful way to own AI. Their two biggest stock positions are Alphabet and Apple. Both fund their AI buildout entirely from operating cash flow, unlike the players borrowing billions and doing circular vendor financing deals for datacenters. Boom continues, BRK participates. Boom pops, the $300B goes shopping.
# Valuation
Not going to discuss ratios here. Gonna keep this part brief to prevent it from getting too quant-y. But two things that stand out from a relative valuation standpoint which I like:
1. BRK.B/SPY ratio is at 0.6618 compared to 1000D-SMA of 0.7933. That's \~19.8% relative upside if the ratio just reverts to its own mean. To be clear, that's relative, not absolute. It can also play out as BRK simply falling less in a drawdown. Either way it's the right shape for what I want this position to do.
2. EPV/MIVoG decomposition at $487.77 ($1.05T cap): Re 11.31% (FF3), EPV $669B, MIVoG 36.4% of price, implied g 4.12%. Financials sector means (n=72): Re 12.87%, MIVoG 50.9%, implied g 7.01%. SPY means (n=511): MIVoG 70.0%, implied g 8.77%. Re-rating to sector mean growth expectations: $631. To SPY mean: $1,035. EPV input is current earnings, \~$300B undeployed in bills, so implied g is biased down relative to deployed-state earnings power.
# Risks
Two that I take seriously. First, clean disinflation. If inflation dies on its own and real rates stay positive, BRK lags and VOO was the better hold. Second, Abel. His operating record is legit but he's unproven as a stock picker, and some of BRK's multiple has always been a Buffett premium.
Feel free to roast/comment. Cheers.