What a Housing Slowdown Does to Each Kind of Trade Work

Why this matters

"Housing is slow" is a headline, not a forecast for your shop. It describes something that will cut one revenue line by a third, leave a second untouched, and possibly push a third up - all inside the same business, in the same quarter. Two shops in the same trade, in the same metro, reading the same news, can come out of the same slowdown with one down by a fifth and the other down by a twentieth, and the only difference between them is where their work comes from. That mix is knowable today and adjustable only before you need it adjusted, which is why this is worth an hour now rather than a panic later.

Five sources, five different clocks

Sort your revenue by where the work originates, not by what the job is. A furnace changeout sold to a builder, to a home inspector's repair list, and to a fifteen-year customer are three different businesses wearing the same invoice.

Two of these rows carry the real content. Renovation is the one everybody reads backwards, and service and repair is the one whose flat revenue line hides a margin change.

Where your trade sits in the build sequence sets your lag

Within new construction, permits are the leading indicator, but the distance from the indicator to your invoice depends on where your trade lands in the build. Site, excavation and foundation work follows a permit within weeks. Rough-in trades - plumbing, electrical, mechanical - arrive once framing is up. Finish trades, final fixtures, and anything tied to a certificate of occupancy arrive at the end.

The practical consequence is that a builder pullback hits an excavator and a finish trade six to nine months apart, off the same permit data. If you are a late-sequence trade, a permit collapse this quarter is a booking problem you will feel next autumn, and the work currently in your backlog is from permits pulled before the market turned. That backlog is the most dangerous comfort in this whole card: it is finite, it is already sold, and it tells you nothing about what comes after it.

If you are an early-sequence trade you get the opposite problem - you feel it first and nobody believes you, including your own suppliers.

The substitution question, which most people get backwards

Renovation does not simply fall in a slowdown. It gets pushed in two directions at once, and which one wins depends on what caused the slowdown.

A rate-driven slowdown supports renovation. When borrowing costs rise, moving gets expensive: the household that would have traded up is now looking at a payment it will not take, so it stays and improves instead. The kitchen, the second bathroom, the system upgrade that would have been the next owner's problem becomes this owner's project. Households that are not moving still have their income. Renovation can hold or rise in a market where sales volume has fallen by a quarter.

An income-driven slowdown does not. When the cause is layoffs rather than rates, the substitution disappears, because substitution requires money the household no longer has. Renovation falls with everything else, and it falls further than service does because almost all of it is deferrable - the split Deferrable Versus Non-Deferrable Work and What a Downturn Touches derives in full.

The tell that separates them is cheap: rate-driven slowdowns show falling sales volume with flat or rising local employment, and income-driven ones show both falling. Read both series, not just the housing one.

Same market, two shops

Two shops in one metro, same trade, both with their revenue sorted by source. The movement figures below are round numbers chosen so the arithmetic is legible; substitute your own local ones.

Shop A, builder-weighted: new construction 55 percent, transaction-linked 15, renovation 10, service and repair 15, agreements 5. That sums to 100.

Shop B, service-weighted: new construction 10 percent, transaction-linked 10, renovation 20, service and repair 45, agreements 15. Also 100.

Now run one rate-driven slowdown across both: permits down 30 percent, closings down 25 percent, renovation up 10 percent on substitution, service and repair flat, agreement renewals down 5 percent.

Shop A: new construction, 55 x 30 percent lost is 16.5 points of total revenue; transaction, 15 x 25 percent is 3.75 points; renovation, 10 x 10 percent gained is plus 1.0 point; service flat at 0; agreements, 5 x 5 percent is 0.25 points. Netting those: 16.5 plus 3.75 minus 1.0 plus 0.25 leaves 19.5 percent of total revenue gone.

Shop B: 10 x 30 percent is 3.0 points; 10 x 25 percent is 2.5 points; 20 x 10 percent gained is plus 2.0 points; service flat at 0; 15 x 5 percent is 0.75 points. Netting: 3.0 plus 2.5 minus 2.0 plus 0.75 leaves 4.25 percent of total revenue gone.

Same market, same trade, same news. Down 19.5 percent against down 4.25 percent, a difference of about 4.6 times, and none of it came from selling better or working harder. It came from where the work originates.

Run the other gate: the same two shops, an income-driven slowdown

Change one input and the answer changes character. Hold permits at down 30 and closings at down 25, but make renovation fall 20 percent instead of rising 10, and agreement renewals fall 15 percent instead of 5.

Shop A: 16.5 plus 3.75 plus (10 x 20 percent) 2.0 plus (5 x 15 percent) 0.75 equals 23.0 percent down.

Shop B: 3.0 plus 2.5 plus (20 x 20 percent) 4.0 plus (15 x 15 percent) 2.25 equals 11.75 percent down.

Shop B's advantage narrows from about 4.6 times to about 2.0 times, and the income run still flatters it, for a reason worth stating plainly. The model holds service and repair flat in both runs, which is fair in a rate shock and too kind in an income one: only the non-deferrable half of service is genuinely income-proof, and the deferrable half, the recommended replacement that becomes a repair or a wait, falls with everything else. Shop B carries 45 points in that line and is charged nothing for it, so 11.75 percent down is a floor rather than a forecast and the 2.0 times is the widest that gap gets. Agreements and renovation carry their income exposure in the arithmetic above; service carries the same exposure and the model suppresses it.

The same flat line hides a second effect. Inside it the mix moves from replacement toward repair, which cuts revenue per call and takes margin down before it takes the top line down at all. So a service-weighted shop meets an income shock twice, in a margin the model never shows and in a top line it understates, which is why it can feel the change a quarter before it can point to it.

What to watch, and what to ignore

Closings, not prices. Transaction-linked work tracks the number of sales. Median price is the number the local news reports and the one that tells you least; a market can hold price while volume halves, and your inspection-repair work halves with it.

Local permit filings, in units, by month. Public, free, and the only leading indicator for new construction. Apply your own sequence lag to it before you react.

Local employment alongside both. This is the single input that tells you which of the two gates above you are in, and therefore what renovation is about to do.

Your own source mix, quarterly. Tag every job with its source at the point of sale. The tag costs a second and it is the only one of these four you cannot buy back later.

Ignore national housing headlines for operational decisions. A national average is a blend of markets moving in opposite directions, and yours is one market. See related: Reading the Leading Indicators That Actually Reach a Trade Shop.

Rebalancing is only available before you need it

Every lever that changes the mix takes one to three years to move a meaningful share of revenue, which means the shop that decides to rebalance during a slowdown is deciding too late to help itself in that slowdown. Build the agreement base, the service reputation and the repeat-customer list while construction is paying the bills, and the cost of doing so is real: those customers are lower-ticket, slower to accumulate, and less satisfying to sell than a builder's schedule.

The honest version of this trade-off is that a builder-weighted shop is not doing anything wrong. It is running a higher-beta business - larger swings in both directions - and the correct response is not to become a service shop but to hold a larger reserve, keep more of its cost variable, and know its own number. A builder-weighted shop with twenty weeks of fixed cost in reserve is in better shape than a service-weighted shop with two. See related: Building a Shop That Survives a Bad Year.

Checking your own read

Print the figure next to the test rather than declaring the read sound:

• Every job carries a source tag, and the five shares sum to 100. Shop A above: 55 plus 15 plus 10 plus 15 plus 5 equals 100. Shop B: 10 plus 10 plus 20 plus 45 plus 15 equals 100. If yours does not close, the untagged remainder is hiding in whichever line you guessed at.
• Shares are of revenue, not job count. New construction and service differ severalfold in ticket size, so a count-based mix will understate builder exposure badly. Both shares above are revenue shares.
• The renovation line has a direction, and you named which gate produced it. In the first run renovation is plus 1.0 and plus 2.0 points; in the second it is minus 2.0 and minus 4.0. If your own forecast moves renovation without naming rate-driven or income-driven, you have guessed.
• Your permit lag matches your position in the build sequence. If you are a finish trade reading permits with no lag applied, your forecast is early by most of a year.

References

• U.S. Census Bureau, Building Permits Survey, for local permit filings in units
• Local multiple listing service or county recorder data, for closings by month rather than median price
• U.S. Bureau of Labor Statistics, Local Area Unemployment Statistics, for the employment series that separates a rate-driven slowdown from an income-driven one
• See related: Deferrable Versus Non-Deferrable Work and What a Downturn Touches (owns the deferrable split), When Rates Rise and Your Customers Stop Financing
• See related: Building a Shop That Survives a Bad Year, Reading the Leading Indicators That Actually Reach a Trade Shop