The First Channel: When Home Equity Stops Being Collateral
For young firms, the first financing route from a housing shock runs through the entrepreneur’s own balance sheet. This chapter focuses on what changes when home equity or personal assets support a business loan, and why the effect is concentrated among owners near a borrowing limit. The goal is to distinguish collateral-enabled entry and survival from the broader business activity that tends to rise and fall with housing demand.
Collateral is borrowing capacity, not just wealth
A house can matter to a business in two different ways. It can make its owner wealthier, which may change spending or willingness to take risks. It can also serve as collateral, giving a lender a claim on an asset if the business loan is not repaid. The second role is the one at issue here. A fall in the home’s market value can reduce the amount the owner is able to pledge, even if the owner’s business prospects and work effort have not changed.
That reduction affects feasible choices. A potential founder may need money before the firm has a revenue history: to pay workers, buy equipment, rent premises, purchase inventory, or cover the gap between delivery and payment. An established young firm may need a working-capital line to keep hiring or to survive a temporary sales interruption. If the owner’s home equity supports the loan, a lower house value can shrink the credit limit or make the loan more expensive. The owner may then delay entry, start at a smaller scale, postpone hiring, or close when a temporary setback becomes difficult to finance.
The key word is “binding.” A price increase does not create useful borrowing capacity for every homeowner. Someone with ample equity already has room to borrow, if they want to. Someone with little equity relative to the mortgage may be unable to borrow against a price gain because legal or lender limits still bind. The marginal value of appreciation is therefore largest for the group close to a financing boundary. The same logic runs in reverse during a decline: an owner with a thin buffer can lose practical borrowing capacity even when the absolute fall in the home’s value looks modest relative to the value of the business.
This is one reason young firms are a natural place to look for the collateral mechanism. They often have little retained profit and few business assets that can be pledged independently. Their owners may be using personal savings or housing equity because the firm itself has not yet established a track record. But age is not a complete explanation. An older local firm with little cash and few alternative lenders may also be constrained, while a young firm backed by substantial resources may be less exposed. Age identifies a group in which thin financial buffers are more likely; it does not reveal the exact constraint for each firm.
The Great Recession evidence is consistent with this mechanism. In the matched business sample, about 9.1 percent of firms in 2008 reported using home-equity finance, and nearly 47.1 percent reported using personal savings or assets. Young firms were more reliant on both: 13.03 percent used home equity and 51.06 percent used personal savings or assets, compared with 7.59 percent and 45.53 percent for old firms. By 2012, the young-firm shares had fallen to 5.35 percent for home equity and 36.12 percent for personal savings, while the old-firm shares were 7.83 percent and 44.19 percent.
The change between 2008 and 2012 should not be read as a clean measure of firms abandoning collateral. The matched survey sample is weighted but covers only part of the main business database, and firms that remain observable after a crisis may be selected. The shift could reflect diminished access to home-equity credit, changes in the firms that survived, or both. Its value is more limited and more useful: it shows that young businesses were notably exposed to personal and home-based finance on the eve of the collapse, making a collateral route plausible and worth testing.
The interaction estimates provide a stronger connection between financing source and sensitivity to prices. Establishments belonging to firms that relied on personal savings were estimated to be roughly twice as sensitive to house prices as establishments at firms using neither home equity nor personal savings. Home-equity reliance was associated with about three times the on-impact sensitivity, and those effects remained persistent, sometimes increasing at later horizons. The comparison is not a random experiment. A firm that uses home equity may be in a sector with different risks, may have fewer conventional financing options, or may be run by an owner who chooses a more volatile growth strategy. The estimates therefore support a channel, not an exact causal decomposition for every firm.
The mechanism also clarifies why collateral is not the same as cash. A house-price fall does not necessarily remove money from a business checking account on the day prices move. It changes the set of future transactions the owner can finance. A hiring plan that depended on a credit line may no longer fit within the lender’s collateral requirement. An investment that would have been profitable at a larger scale may be postponed because the owner cannot bridge the initial cost. The effect appears in entry, employment, and survival because financing determines which choices are feasible before returns are realized.
What the housing-boom comparison reveals
The earlier Census microdata study helps separate this capacity mechanism from the broader opportunities created by a housing boom. Across the Texas region during 1992–2007, a 10 percent increase in house prices was associated with about a 0.21 percentage-point increase in the local entry rate, whose average was approximately 8.82 percent. That relationship could reflect rising demand for construction, real-estate services, and other local activities. It could also reflect more available collateral. The correlation alone cannot say how much came from either source.
Texas supplied a more focused institutional change. Before 1998, the state’s constitutional restrictions largely prevented homeowners from using home equity to finance consumption or business investment beyond the home. A 1998 amendment permitted home-equity loans up to an 80 percent loan-to-value ratio, inclusive of the primary mortgage, without restricting how the funds were used. A further change in 2003 allowed home-equity lines of credit. Because these changes expanded access to housing-backed borrowing at identifiable dates, they offered a way to look for entry that was specifically associated with newly available collateral rather than with house prices alone.
The result was positive but modest. Post-reform Texas zip codes experienced additional entry when prices rose, especially after HELOCs became available. In one specification, the extra response was about 4.1 percent of the broad house-price/entry correlation; across specifications, the estimated collateral effect ranged roughly from 3 to 8 percent. The study’s overall conclusion was that the collateral channel accounted for a tenth or less of the entry associated with house-price increases.
That estimate is easy to misread. It does not say that collateral never matters. It says that most people who enter during a housing boom are not entering solely because a rise in home value relaxed a previously binding credit constraint. Housing booms create opportunities of their own. Consumers spend differently, construction expands, and businesses that serve a growing local market may enter because expected demand is higher. A home-equity loan can be a useful financing instrument without being the event that made entry possible.
The composition of entry reinforces this distinction. The broad response linked to house-price growth was concentrated in construction and real estate, and about 82 percent of the entry associated with the 2003–2007 price gains came from firms entering with one or two employees. These entrants tended to be small and short-lived. Industries tied to construction and real estate accounted for about 65 percent of the typical entry response. By contrast, the excess Texas entry associated with the lending reform was more balanced across sectors, more likely to involve larger entrants, and more likely to survive five or more years. Only about a quarter of the post-reform response came from the construction and real-estate-related group.
The comparison is informative without proving that every construction entrant was demand-driven or every reform-linked entrant was high quality. It shows why a large aggregate relationship can contain a smaller financing component with different firm characteristics. Opportunity-driven entry may expand and disappear with the local housing cycle. Collateral-enabled entry may be less common but support firms with a broader sector mix and longer survival. The mechanism changes the type of entry, not just the number.
The individual-level evidence makes the borrowing boundary even clearer. In the study’s homeowner sample, about 85 percent were classified as unlevered: they had accumulated enough equity before the 2000–2004 boom that further price appreciation did not materially change their borrowing capacity. About 13 percent were classified as levered, and about 2 percent as lending-constrained because legal limits or local price declines prevented them from using new equity. Once wealth, demographics, earnings, and related traits were controlled for, the large unlevered group showed little additional entry response to house-price gains.
The strongest response came from the smaller levered group that experienced large gains. For the roughly 3–6 percent of wage workers most plausibly affected by both high leverage and substantial price appreciation, entrepreneurial transition rose by about 50 percent relative to the sample-average entry rate in the reported specification. A large effect for that group can coexist with a small aggregate collateral share because the group is small. This is the central distributional lesson of collateral: its average contribution may be limited while its marginal value is very high for an owner close to the constraint.
Survey evidence points in the same direction. In the 2007 Survey of Business Owners, about 12 percent of new employer businesses reported using home equity for startup finance, compared with about 16 percent using bank loans and 18 percent using credit cards; personal savings and assets were more common. A 10 percent state-level house-price increase from 2000 to 2007 was associated with only a 0.3 percentage-point increase in the share of entrants using home-equity finance. Respondents could report more than one source, and the survey does not show how much of the startup budget each source provided. These figures indicate meaningful use, not a large population-wide unlocking of blocked firms.
Taken together, the evidence gives the collateral channel a precise role. It is not a universal wealth-to-entrepreneurship law, and it is not sufficient to explain all firm entry during a housing boom. It is a financing capacity mechanism that becomes powerful for owners with thin equity buffers, especially when a young firm has few alternative assets or lenders. That concentration is exactly why aggregate averages can understate the consequences for a particular firm.
The contrast also prepares the broader two-channel comparison. The earlier study follows rising prices and firm entry, while the Great Recession evidence follows a collapse in prices, continuing employment, entry, exit, and local bank exposure. A collateral effect that is a small share of boom-era entry can still be consequential when a downturn removes a young firm’s last buffer. To understand the rest of the employment decline, however, we must examine the balance sheet on the other side of the loan: the local bank that may be losing capacity at the same time.