Our Perspective
Q2 2026
Farnam Street Wisdom
The price of being approximately right is going up.
There are periods in business when a rising tide forgives a great many sins. Capital is cheap, customers are accommodating, inventories eventually clear, mediocre investments can be refinanced, and growth has a pleasant habit of making yesterday’s mistakes look like tomorrow’s strategy.
This does not appear to be one of those periods.
The economy is still growing. Consumers are still spending. Factories are still producing. Capital is still available. Deals are still getting done. None of those statements is particularly controversial.
The mistake would be assuming that because the machine is running, it is running easily.
It isn’t.
The interesting feature of the current environment is not weakness. It is selectivity.
Consumers are becoming more selective about what they buy and where they buy it. Retailers are becoming more selective about what earns shelf space. Lenders are becoming more selective about whom they finance. Investors are becoming more selective about which growth stories deserve another round of capital. Buyers are becoming more selective about which businesses deserve strategic premiums.
Even the economy itself seems increasingly selective about which mistakes it is willing to forgive.
That makes this a particularly good time to remember a few old lessons.
Revenue is not economics.
There may be no more expensive sentence in business than, “But sales are growing.”
Growth is useful when each additional dollar of revenue leaves the business stronger. It is considerably less useful when it requires two dollars of inventory, another promotion, longer payment terms, additional headcount and a larger revolver.
This distinction is especially important in consumer goods.
A new retailer can produce an impressive purchase order and an equally impressive working-capital problem. Distribution can increase while velocity deteriorates. Gross revenue can rise while trade spending quietly consumes the improvement. A brand can be “everywhere” shortly before discovering that everywhere is an expensive place to be.
The shelf does not care about the founder’s valuation. Neither does the consumer. Eventually, somebody has to buy the product twice.
That is why we increasingly prefer boring evidence to exciting narratives: velocity, repeat purchase, contribution margin, cash conversion, inventory turns and returns on incremental capital.
A business that understands those numbers has options. A business that does not is usually financing an education.
Capital has not disappeared. It has acquired standards.
There is a peculiar contradiction in today’s capital markets. Enormous amounts of money are being invested while many perfectly respectable businesses find raising capital considerably more difficult.
There is nothing contradictory about it. Capital does not distribute itself democratically.
When investors become uncertain, money tends to congregate around things that appear unusually obvious: scale, scarcity, cash flow, strategic importance or the possibility of extraordinary returns.
The result is a market that can look euphoric from 30,000 feet and rather stingy from the parking lot.
This is particularly visible in venture capital. Record amounts of money can coexist quite comfortably with difficult fundraising conditions for the median company.
The average, as usual, is capable of telling the truth while giving entirely the wrong impression. For operators, the lesson is straightforward: never confuse the existence of capital with your entitlement to it.
The best financing strategy remains building a company that does not need to accept the first financing offered. Cash creates time. Time creates choices. Choices improve negotiating leverage. This is not complicated finance, but it is complicated behavior.
Debt does not become equity because the interest-only period is generous.
When equity becomes expensive, people develop a remarkable ability to discover the virtues of debt. Debt has many virtues. Patience is generally not among them.
Used against receivables, equipment, contracted demand or a predictable working-capital cycle, debt can be an excellent instrument. Used to postpone discovering whether the underlying business works, it is merely a deadline with covenants.
We occasionally hear debt described as “non-dilutive capital.”
That is true in roughly the same sense that rent is non-dilutive housing.
The lender may not own the upside, but he remains surprisingly interested in being paid.
The correct question is not whether debt or equity is cheaper. It is what uncertainty the capital is being asked to finance.
Equity is unusually good at absorbing uncertainty.
Debt is unusually good at financing things that are expected to happen.
Confusing the two is how temporary optimism becomes permanent leverage.
The consumer has not disappeared. The lazy purchase has.
There is a temptation whenever consumers become cautious to describe them as “weak.”
That description misses what is actually happening.
The American consumer continues to spend. But households are increasingly sorting purchases according to value, convenience, health, quality, experience and necessity.
They are shopping across channels with considerably less loyalty to the old boundaries between premium, mass, club, discount and private label.
A wealthy household shopping at Costco is not necessarily trading down. It may simply dislike wasting money.
A lower-income household buying one premium product is not necessarily behaving irrationally. That product may matter enough to justify economizing somewhere else.
Consumers do not carry economists’ spreadsheets into grocery stores. They make trade-offs.
Businesses that understand those trade-offs are gaining share.
Those that rely on the customer behaving the way she behaved five years ago are conducting a rather expensive sociology experiment.
This matters because price is only one form of value.
Convenience is value. Reliability is value. Portion size is value. Health is value. Time is value. Trust is value.
The winning company does not necessarily offer the cheapest product.
It makes the purchasing decision easiest to defend.
A wonderful technology can still be a terrible investment.
Perhaps the most useful historical reminder this quarter comes from the railroads.
Railroads transformed America. They connected farms to cities, manufacturers to national markets and consumers to goods they could never previously obtain. They were unquestionably important. They also bankrupted an impressive number of investors. There is no paradox.
A technology can create enormous value for society while competition, excessive capacity, leverage and enthusiastic financing transfer much of that value away from the people who funded it. The internet did something similar. Artificial intelligence may as well.
We have little doubt that AI will alter how companies operate. It will make certain forms of analysis cheaper, automate repetitive work, improve forecasting, accelerate software development and make information dramatically easier to manipulate.
That tells us remarkably little about which AI investment will produce an attractive return.
Usefulness and investment merit are cousins, not twins.
When everybody agrees that something will change the world, the interesting question is usually no longer whether they are right.
It is what price they are paying to be right.
Good Businesses Should Be Getting More Boring
Business has a peculiar tendency to make simple things complicated when times are good.
More SKUs look like growth. More customers look like scale. More debt looks like efficient capital. More initiatives look like progress. Eventually, a company can become very busy without becoming much better.
Less forgiving markets tend to correct that.
The good operators start asking simpler questions. Which products actually make money? Which customers are worth serving? Where is the cash getting stuck? What are we doing simply because we have always done it?
The answers are rarely glamorous. Carry less inventory. Collect faster. Drop the SKU that never earned its keep. Reprice the customer that consumes more profit than it produces. Keep a little more cash than seems necessary.
None of this will make the company look terribly sophisticated.
That is probably fine.
Complexity should have to earn its way onto the income statement. If it does not improve the economics of the business, it is usually just another thing to manage.
The best businesses are not built around correctly predicting next year. They are built so they don’t have to.
Boring businesses tend to survive exciting times.
The AmpliFi View
We do not see an economy falling off a cliff. We also see little reason to behave as though the road ahead is straight.
Demand remains. Capital remains. Opportunity remains. What has become scarcer is forgiveness.
When money was cheap and growth plentiful, a surprising number of mistakes could be refinanced, repriced or simply outgrown. That luxury is fading. Businesses increasingly have to earn their economics rather than explain them.
We find that encouraging.
Periods like this tend to favor companies that understand why they make money, where they consume cash and what could make them wrong. The advantage belongs less to the company with the most confident forecast than to the one that can withstand the forecast being wrong.
That is margin of safety in an operating business.
And right now, we would take a little more of it over another decimal point in the forecast.
Contrarian Insight | There Is No Average Customer
The economy grew 1.5% in the second quarter.
It is a useful number. It is also a number that no business actually experiences.
No customer walks into a store and spends 1.5% more because GDP increased. No farmer receives the national average commodity price. No manufacturer borrows at the federal funds rate. No trucking company hauls the average freight load. And very few businesses sell to the average American consumer.
Averages become most dangerous when the things being averaged are moving apart.
That appears to be happening now.
There is an American economy in which hundreds of billions of dollars are being committed to artificial intelligence, data centers, power generation and the infrastructure required to support them. Land is being acquired, electricity contracted, equipment ordered and people hired at a pace that looks considerably more like a boom than a 1.5% economy.
There is another economy in which a farmer is borrowing near 7% against tighter crop margins and wondering whether another piece of equipment can make it one more season.
Both are America.
There is a consumer who can absorb higher prices, continue traveling, shop at Costco for value and pay a premium for something that saves time or improves health.
There is another consumer comparing unit prices, moving purchases to private label, eating out less often and deciding which expenses can wait until payday.
Both show up in retail sales.
There is a large corporation issuing debt into a public market where spreads look remarkably calm.
There is a smaller business sitting across from a commercial banker being asked for another covenant, another guarantee and a little more equity.
Both are supposedly borrowing in the same credit market.
This is the problem with averages. They are mathematically correct and occasionally economically useless.
For a long time, businesses could get away with managing to them. If the consumer was “strong,” add inventory. If GDP was growing, increase the sales forecast. If rates were coming down, assume financing would get cheaper. If the category was expanding, add capacity.
That worked reasonably well when the tide was moving most boats in roughly the same direction.
It works considerably less well when the boats are heading different ways.
The important question for an operator today is therefore not whether the American economy is strong or weak.
It is which economy are you actually in?
If your customer is a middle-income household buying groceries, the answer may be hiding in basket composition, private-label penetration and promotional response—not GDP.
If you manufacture food, your economy may be determined more by cocoa, milk, packaging, freight and retailer deductions than by consumer confidence.
If you farm, national employment statistics are considerably less useful than local basis, land rents, input costs, crop prices and the attitude of your lender.
If you operate a factory, 1.5% GDP matters considerably less than what is happening to your order book, utilization, labor availability and the price your suppliers are quoting for the next six months.
This sounds obvious.
Most useful things do after somebody says them.
The danger comes when management uses the macroeconomic story to explain away evidence sitting directly in front of it.
“The consumer is resilient” can become an excuse for weak velocity.
“The economy is slowing” can become an excuse for losing share.
“Inflation is moderating” is not particularly comforting when your own landed costs are rising.
And “the category is growing” is of limited consolation if someone else is doing all the growing.
The national economy matters. Interest rates matter. Inflation matters. GDP matters.
But they are the weather map, not the view through the windshield.
At 1.5% growth, America is not uniformly booming and it is not uniformly struggling. It is sorting—by income, industry, geography, access to capital, competitive position and increasingly by the ability of individual companies to translate demand into cash.
That sorting creates winners that look strange against the macro numbers and losers with perfectly respectable explanations for why they lost.
For operators, the lesson is less exciting than predicting the next recession and considerably more valuable:
Know your customer better than the consumer.
Know your costs better than the CPI.
Know your demand better than GDP.
And know your business well enough that when the averages tell you one thing and your economics tell you another, you know which one to believe.
The average economy is an interesting place to study.
Just don’t make the mistake of thinking your company operates there.
