The July retail sales report lands at 8:30 this morning, and it will be reported as one number describing one consumer. There is no such consumer. There are at least two, they're moving in opposite directions, and the published data is detailed enough to tell you which one you are.
Start with what the aggregate actually says, because it's genuinely strong and I don't want to talk you out of the facts. June retail and food services sales came in at $768.6 billion, up 0.2 percent on the month and 6.7 percent from a year earlier, per the Census Bureau. The soft-looking monthly figure was almost entirely a gasoline artifact: receipts at gas stations fell 5.3 percent because pump prices dropped, not because anyone bought less fuel. Strip gas out and sales rose 0.7 percent. The control group, which excludes autos, gas, building materials, and food services and feeds directly into the GDP calculation, rose 0.5 percent. Nonstore retail, mostly e-commerce, climbed 1.9 percent on the month and was running more than fourteen percent above the prior year.
Set that 6.7 percent annual growth against consumer prices running at 3.5 percent and you get roughly three percent real growth in consumption. That is not a struggling consumer. On the surface, it's a strong one.
Now go one layer down, and the picture stops being a picture of one economy.
Total household debt reached a record $18.8 trillion in the first quarter of this year according to the New York Fed's quarterly report on household debt and credit, which is built from a nationally representative sample of anonymized credit files rather than a survey. Mortgage balances stood at $13.19 trillion. Credit card balances were $1.25 trillion, down seasonally from the fourth quarter but up roughly six percent from a year earlier. Auto loans reached $1.69 trillion. Student debt sat at $1.66 trillion, with 10.3 percent of balances ninety or more days delinquent. Roughly 4.8 percent of all outstanding household debt was in some stage of delinquency, and about 124,000 consumers picked up a bankruptcy notation in a single quarter.
The line I'd underline, though, is a smaller one. Home equity line balances rose again to $446 billion, marking the sixteenth consecutive quarterly increase, and sitting roughly $129 billion above the low reached in early 2022. Sixteen straight quarters is not a blip. It's a behavior.
The New York Fed's own researchers characterized household credit as broadly stable while noting weakness concentrated in lower-income households, and mortgage delinquency deterioration concentrated in lower-income areas and places where home prices have been falling. That framing is precise and it's the thing worth carrying with you: the aggregate is fine, and the aggregate is hiding a distribution.
Put the two data sets next to each other and the mechanism becomes visible. Spending has held up. But a growing share of it is not being funded by income. It's being funded by two different things depending on where you sit.
At the top, it's being funded by asset appreciation. Analysts covering the June report noted directly that spending growth continues to be driven by higher-income households whose balance sheets have been inflated by the equity rally. That's the wealth effect, and it's real and well documented: people spend more when their portfolios are up, even though nothing was sold and no income was earned. The problem is what it does to the structure of the economy. When consumption depends on asset prices, and corporate earnings depend on consumption, and asset prices depend on earnings, you've built a loop that reinforces itself in both directions. It feels like strength on the way up. It's reflexivity, and reflexive systems don't decay gently, they reverse.
In the middle, it's being funded by credit extension, and the HELOC number is the cleanest evidence of it. Think about who takes a home equity line in 2026. Overwhelmingly, it's someone sitting on a mortgage locked at three or four percent who cannot sell without giving up that rate, has substantial paper equity from five years of price appreciation, and needs cash. They can't refinance the whole mortgage without repricing the entire balance at today's rates, so instead they draw a second lien on top. Sixteen consecutive quarters of that is a country converting illiquid, locked-in home equity into current spending, one line of credit at a time.
And here's the piece that connects to everything I wrote on Monday. Home equity lines are floating rate. So are credit card balances. So are most business revolvers. They reprice off short-term benchmarks within a billing cycle or two of any move by the Federal Reserve. With market-implied odds recently favoring a hike at the September meeting rather than a cut, the exact channel that has been financing consumption at the margin is the channel that gets squeezed first. Not eventually. Within about sixty days of the decision.
There's also a category of obligation that doesn't appear in any of these figures at all. Buy now, pay later arrangements have expanded into a substantial pool of household commitments over the past few years, and most of them sit outside traditional credit reporting entirely, which means they don't show up in the New York Fed's panel, don't affect credit scores in the usual way, and don't get counted when anyone tallies household leverage. Estimates of the size vary widely and I'd treat any specific figure with caution, but the direction isn't in dispute: the reported debt numbers are a floor, not a ceiling, and the gap is concentrated in exactly the population where the balance sheet is thinnest. When a category of borrowing grows fastest precisely where the measurement is weakest, the aggregate statistics get less useful at the moment you most need them.
And several of the cushions that supported spending through the first half of the year were one-time in character rather than structural. Unusually generous tax refunds provided a cash injection concentrated in the spring. Households have been drawing on savings accumulated in prior years. Falling pump prices through June freed up cash that flowed straight into other categories, which is a large part of why non-gasoline spending looked so healthy in the June report. Each of those is real money and each of them is finite. A consumer supported by a refund, a savings drawdown, and cheaper gas is not the same thing as a consumer supported by rising real wages, even when the spending line looks identical on a chart.
So the useful question this morning is not whether the consumer is strong. That question has no answer, because the consumer isn't a person. The useful question is which side of the distribution you're on, and that one is answerable in about twenty minutes.
Three diagnostics, in order of how much they tell you.
The first is the funding source of your own last twelve months. Take your total spending across the past year and account for where every dollar of it came from. Income after tax is one bucket. Net new borrowing, meaning any increase in credit card balances carried past the statement date, any HELOC draw, any new installment loan for something you consumed, is the second. Drawdown of savings or sale of assets is the third. Almost nobody has ever done this arithmetic, and it's brutally clarifying. If bucket one covered everything and left a surplus, you're on one side of the line. If buckets two or three did meaningful work, you're on the other, and the fact that your net worth may have risen over the same period because your portfolio went up does not change that. Rising net worth alongside rising debt-funded consumption is precisely the pattern that looks healthy right up until asset prices stop cooperating.
The second diagnostic is your revolving balance behavior, not your balance. Plenty of people run substantial charges through a card every month and pay it in full, and their statement balance tells you nothing. What matters is whether there was any month in the last twelve where interest was actually charged. One month is a signal. Three or more is a structural condition, and it means your effective cost of capital on marginal spending is somewhere in the high teens or low twenties, which is a rate no investment portfolio in your lifetime is going to outrun.
The third is liquid reserves measured in months of actual expenses, not in dollars. Dollars are a number that sounds reassuring. Months are a number that tells you how long you have. The reason this one matters most in the current environment is that it determines whether a disruption becomes an inconvenience or becomes a HELOC draw at a floating rate in the middle of a possible hiking cycle. Reserves are what keep you from being forced onto the wrong side of the distribution by a single bad quarter.
Assembling the inputs for all three is the part that stops people, since the answer requires income, spending, and every liability in one place at the same time. Pulling it into a single view with a tool like Empower is what turns this from a weekend project into a coffee-length one, and the specific thing you want on screen is the change in total liabilities over twelve months sitting directly beside total spending over the same twelve months. Those two numbers together answer the funding question faster than any budget ever will.
If you run a business, this distribution is not an abstraction, it's your revenue forecast. Your customer base sits somewhere on this line, and the two halves behave completely differently under stress. Customers funded by asset appreciation cut discretionary spending quickly when markets fall but have deep reserves. Customers funded by credit extension keep spending longer than they should and then stop abruptly and completely, often with a chargeback attached. If your revenue is concentrated in the second group, your genuine risk isn't a recession, it's a hundred basis points on short rates. Knowing your mix is worth more than any demand forecast you'll buy, and the operational hedge is the same one it always is: fixed costs converted to variable where possible, and recurring manual work moved onto an automation layer like Make so your cost base can flex without your headcount having to.
On the personal side, the response to being on the credit-funded side of the line is unglamorous and completely reliable. Rebuild reserves before optimizing anything else, because reserves are what buy you the option to make good decisions later. Convert what floating-rate exposure you can convert to fixed while the option is still cheap. And keep the investing contributions running on a schedule rather than on sentiment, which is what routing the allocation layer through a platform like M1 Finance is actually for, because the single worst outcome available here is stopping contributions during the drawdown and restarting after the recovery.
The response to being on the asset-funded side is different and slightly counterintuitive. Your risk isn't insolvency, it's that your spending has quietly indexed itself to a portfolio value that can fall thirty percent without anything unusual happening. The discipline there is to size your fixed commitments to your income rather than your net worth, which is a rule wealthy households violate constantly and then discover during exactly the wrong quarter.
Whatever the number is at 8:30, watch the gasoline line and the control group rather than the headline, and watch nonstore versus food and beverage stores. Discretionary strength alongside softness in staples tells you the top is carrying the number. That's the tell, and it will be in the report.
I built the tracker I use as a one-page dashboard: the six series that actually describe consumer health, where to find each one for free, the threshold on each that separates normal from deteriorating, and the three personal diagnostics above with the arithmetic already laid out. Reply with the word PULSE and I'll send you The Consumer Signal Dashboard. It takes twenty minutes to fill in and it tells you which economy you're living in.
If this is the kind of analysis you want more of, forward it to someone who'd use it. Refer three people to Money Systems Lab and the full playbook library unlocks. Refer ten and you get lifetime premium access, including the complete Wealth Architecture Blueprint, our course on building the entire system: portfolio construction, cash management, tax positioning, and the automation layer that keeps it running without you.
On Sunday I'm going to bring this week together around the single question that determines whether a rising-rate environment helps you or hurts you, and it isn't the one most people ask.
The aggregate is an average of people having completely different experiences. Find out which one you're having, in numbers, before the market tells you.
Taylor Voss
Money Systems Lab
Institutional-grade financial intelligence for everyone else.
Disclosure: Some of the links above are affiliate links, which means Money Systems Lab may earn a commission at no additional cost to you if you choose to sign up. I only mention tools I actually think fit the systems approach I write about. None of this is personalized investment, tax, or legal advice. Do your own research and weigh your own situation before making any financial decision.