Diagnosing the Financing Shock

The central subject of this series is the young and small U.S. firm exposed to a housing shock. The complete answer to why housing declines hit these firms so hard is that one market movement can damage two different balance sheets: it reduces the collateral an entrepreneur can pledge and can weaken the local bank that supplies credit. The two channels meet in entry, employment, and survival, but they do not operate on the same timetable or leave the same evidence. A sound diagnosis therefore asks which balance sheet is binding, which firm margin is moving, and how much confidence the evidence deserves before recommending a response.

One housing shock, two financing routes, one firm lifecycle

The borrower-side route begins with pledgeable assets. A homeowner who uses home equity or personal assets to support a business may lose borrowing capacity when house prices fall. The effect is largest near a constraint: an owner with ample equity may remain able to finance a plan, while a highly leveraged owner may have to abandon entry, reduce hiring, or accept a smaller scale. The Great Recession evidence is consistent with this distributional pattern. Young firms were more reliant on home-equity and personal-asset financing in 2008, and firms reporting those sources were more sensitive to house prices. The indicators show exposure to a plausible channel, not a random assignment or a complete record of every loan decision.

The lender-side route begins elsewhere. A local bank with a housing-exposed balance sheet can face lower net worth, greater leverage pressure, and a higher lending spread when house prices fall. A small business can then be constrained even if its owner has not pledged a home. The evidence links pre-crisis mortgage shares in state-focused banks to a stronger house-price employment response for young and old/small firms. The measured interaction is a proxy for local credit-supply exposure, not an observation of a bank refusing a particular loan. Its importance depends on geographic segmentation: firms that rely on nearby relationship lenders have fewer immediate substitutes when those lenders tighten.

The channels are distinct but can reinforce each other. A young firm with a thin home-equity buffer may encounter a bank demanding more protection at precisely the moment that the owner has less to pledge. Another firm may have ample owner collateral but still face a higher rate or reduced loan limit because its local bank is under pressure. A third firm may face neither financial constraint and respond mainly to local demand. Treating all three cases as one “housing effect” would hide the mechanism that matters for both diagnosis and policy.

The firm lifecycle determines how the pressure appears. Before entry, collateral or credit supply can prevent a business from opening at all. After entry, the firm may cut employment or investment while trying to continue. If the pressure persists, it may exit; once it has exited, a later recovery does not automatically restore the lost firm or its jobs. An older small firm can remain exposed to local bank credit even after it has accumulated experience, while an old/large firm may be less dependent on any one local lender. Age is therefore a marker of different financial positions and alternatives, not a complete causal explanation.

This lifecycle is why the empirical pattern has several dimensions. Instrumented estimates associate a 1 percent increase in metro house prices with about a 0.289 percentage-point increase in young-firm employment growth, compared with about 0.142 points for old/small firms and 0.0505 points for old/large firms. A 25 percent national decline is translated into roughly 5.4–6.4 percentage points of young-firm employment decline and about 3.3 points for old/small firms. But employment growth is not the whole outcome. Excluding entrants removes much of the young-firm excess effect; exit probabilities also rise, by about 0.05 percentage points for young firms and 0.025 points for old firms for a 1 percent house-price decline.

Those estimates describe a connected time path rather than separate losses to be added mechanically. Young firms are hit more strongly at first, but their excess sensitivity relative to old/small firms fades after roughly two later periods while the overall effect persists. Local bank interactions also fade faster than the total house-price effect.