A private individual, most likely. Seventy percent of US mortgage dollars in the early 1890s came from individual investors rather than institutions, in part because banks and insurers were barred by statute from lending across state lines.
In the early 1890s, total mortgage debt outstanding in the United States came to roughly $6 billion. That figure covered every loan secured by real estate that had not yet been repaid: homes, farms, and commercial property, held by lenders of every kind.
Seventy percent of those dollars came from individual investors. Commercial banks, mortgage companies, savings and loans, building and loan associations and insurance companies together accounted for the other thirty. Insurance companies, the third largest intermediated source, held about $400 million, roughly 7 percent of the market.
The individuals were not a uniform group. Some were professional operators and real estate attorneys who made loans as a business. Others were sellers carrying paper for the buyers of their own homes, family members, or occasional investors placing savings they happened to hold.
The absence of institutions was not only a matter of preference. Much of it was statutory.
The National Bank Act of 1864 prohibited federally chartered banks from lending secured by real estate. That prohibition stood until 1913. New York life insurance companies, which held half of all US life insurance assets, could invest only in mortgages within the state or within fifty miles of New York City until at least the late 1870s, and were barred from lending outside New York entirely until 1886. Mutual savings banks were typically confined to their home states. New York building and loan associations could not lend on property more than fifty miles from their own headquarters.
Several other states imposed similar limits on insurers chartered within them, though some large companies in Connecticut and Wisconsin held interstate lending powers. The effect was that the country's deepest pools of capital were legally fenced off from the markets with the greatest demand.
An individual lender was, by definition, poorly diversified. Only 24 percent of individual investor funds came from outside the borrower's state. A lender typically held a small number of loans concentrated in one or a few localities, so a single default took a large proportion of the position, and the interest rate had to carry that risk.
Concentration was geographic as well as numerical. In 1890, 42 percent of the population lived on farms, and a lender to farmers in one county faced correlated risk: an unexpected freeze or a bad harvest moved every loan in the portfolio in the same direction and in the same season. In industrial areas, the failure of a single significant employer produced the same effect.
Because savers were not evenly distributed and lending was overwhelmingly local, interest rates varied considerably by region. Where savings were concentrated, in the Northeast, lenders competed with one another and terms improved: lower rates, longer maturities, smaller down payments. Where savers were fewer, in the West and South, rates ran higher.
The consequence was allocative rather than merely unfair. More homes were financed in high-savings regions even where demand was greater elsewhere, and households entirely able to carry a loan went unserved because capital would not travel to them.
Whether the regional spreads were justified by risk remains disputed among economic historians. Davis (1965) argues a national mortgage market had largely developed by 1900, though rates stayed unusually high in the South and to some extent the West. Eichengreen (1984) concludes the East-West and East-South differentials are mostly explained by differences in foreclosure risk. Snowden (1987), examining both farm and home mortgage rates in 1890, disagrees and finds regional differences that survive adjustment for default risk. The question is not settled.
The binding constraint was monitoring. A lender needed to understand local business conditions well enough to forecast land values, and to confirm the borrower was maintaining the property that secured the loan. A local lender could do this cheaply: he knew the borrower, the block, and the building.
Lending at a distance meant either travelling to inspect collateral or hiring someone to do it, and then finding a way to ensure the delegated monitor was diligent and honest. Price and Walter note that costly mistakes arose in the nineteenth century when delegated monitors were careless or dishonest, in the same way they did during the subprime crisis of the twenty-first.
Even after legal restrictions were lifted, the economics often held. New York insurers lobbied hard through the 1870s and 1880s to be allowed to lend out of state, having watched Connecticut companies do so profitably. When the restrictions were removed, the cost of lending in distant markets was frequently found to be prohibitive anyway.
In 1890 the limit on lending was physical. A lender financed what he could travel to see, and the map decided who obtained credit and at what price.
The constraint is no longer physical, but the question it created has never gone away. Every structure built since, from the mortgage companies of the 1870s through the government sponsored enterprises, has been an attempt to lend far from the collateral and still understand what is held. Diversification and scale pull in one direction; local knowledge and effective monitoring pull in the other. That tension is the subject of the paper this issue draws on, and it is unresolved.
Geography still sets the price of a mortgage. Measuring it now means observing every originator in every market, at loan level, down to the census tract, across all 3,200 counties.
A note on what is contested. Whether mortgage rates in the West and South were higher than borrower risk justified is an open question in the literature. Eichengreen attributes the spread largely to foreclosure risk; Snowden finds differences that persist after adjusting for it. The carousel states that rates were higher and that capital did not travel. It does not claim the spread was unjustified.
