Special Report

Housing Finance and the Government-Sponsored Enterprises

Of everything on this site, the housing-finance system is the subject where the gap between how often it is mentioned and how rarely it is explained is widest. It is also the clearest case study economics offers in what happens when an institution is placed deliberately between the public and private sectors and left there for decades.

The problem the system was built to solve

A mortgage is an awkward asset. It is very long-dated, it is illiquid, its risk depends on a local property market, and the borrower can usually repay early — which means the lender's asset disappears precisely when interest rates make it most valuable to keep. A bank funding thirty-year mortgages from deposits that can be withdrawn on demand is running a serious maturity mismatch, and history is full of institutions that discovered this the hard way.

The response, developed over the twentieth century in the United States and copied in various forms elsewhere, was securitisation: pool many mortgages, sell claims on the pooled cash flows to investors, and thereby move the funding of housing out of the banking system and into the capital markets. Done well, this genuinely lowers borrowing costs — it broadens the pool of lenders, diversifies idiosyncratic risk, and lets long-term investors such as pension funds hold long-term assets that suit them.

What a government-sponsored enterprise is

To make that market work, the United States chartered a particular kind of institution: a private, shareholder-owned company created by statute to pursue a public purpose, with a set of legal privileges attached. The two large housing enterprises bought mortgages from originators, guaranteed the credit risk on the securities they issued, and held a large portfolio of mortgage assets on their own balance sheets.

The privileges mattered more than they appeared to. A line of credit with the Treasury, exemption from certain taxes and registration requirements, and a statutory charter combined to create a widespread market belief that the government would not let these institutions fail. That belief was never a legal guarantee — every security carried a printed statement that it was not backed by the government — and every serious participant nonetheless priced as though it were.

Why an implicit guarantee is a problem

The economics of this arrangement are worth stating carefully, because it is a general lesson rather than a story about two companies.

An implicit guarantee lowers an institution's funding costs. That subsidy is real and is captured partly by borrowers in cheaper mortgages and partly by shareholders in higher returns; the split has been estimated many times, with a wide range of results. Crucially, the subsidy grows with the size of the balance sheet, which gives management a strong incentive to expand — and because the downside is socialised while the upside accrues to shareholders, the incentive to take risk is stronger than for an ordinary company. Economists call this a moral hazard, and it is the standard diagnosis here.

It is also difficult to price. The subsidy does not appear in any budget, no appropriation is voted, and the contingent liability sits off the public balance sheet until it is realised. That opacity is not incidental; it is the feature that made the arrangement politically durable for decades.

The accounting restatements

In the early and mid 2000s both enterprises restated years of financial results, in one case by an amount in the billions. The technical issue centred on hedge accounting: rules that permit a firm hedging an interest-rate exposure to defer recognition of gains and losses, provided the hedge is documented and demonstrably effective. Regulators concluded the documentation and testing standards had not been met, and that earnings had accordingly been smoothed across periods in ways the rules did not allow.

Two features made this more than a technical dispute. Earnings smoothing had a bearing on executive compensation, which was tied to earnings targets. And the episode revealed that institutions of enormous systemic importance had internal controls and regulatory oversight considerably weaker than their scale warranted — a supervisory gap that had been noted by outside analysts well before it became a crisis.

Conservatorship

In September 2008, with mortgage credit losses mounting and funding markets seizing, both enterprises were placed into conservatorship under a newly created federal regulator, with Treasury support agreements putting public capital behind them. The implicit guarantee became explicit. Whatever one concludes about the wisdom of the decision, it settled a long-running empirical question: the market's belief that these obligations were effectively public had been correct.

The enterprises subsequently returned to profitability and have paid very large sums to the Treasury. They also remain in conservatorship many years later, which is itself informative — the difficulty of designing a successor system that preserves the long fixed-rate mortgage without recreating the same implicit guarantee has defeated repeated legislative attempts. The Federal Housing Finance Agency publishes the current regulatory position and the historical record; the Congressional Budget Office has published repeated analyses of the fiscal treatment and the reform options.

What the episode actually teaches

Three durable lessons, none of them partisan.

• Ambiguous public-private status is unstable. An institution that is private in profit and public in loss will, over a long enough period, be tested on that arrangement. The failure mode is predictable in advance and was in fact predicted.
• Contingent liabilities that appear nowhere still exist. A guarantee that does not require an appropriation is not free; it is unmeasured. Modern budget scoring has moved somewhat toward recognising the fair value of such commitments precisely because of this history.
• Accounting rules are risk controls. The hedge-accounting requirements that were breached exist to prevent exactly the information loss that occurred. Treating disclosure standards as paperwork is how supervisory failures begin.

Reading housing-finance coverage

Distinguish the credit question (will borrowers repay) from the rate question (what happens to the value of the assets when rates move) — they are different risks with different remedies and are constantly conflated. Distinguish a guarantee fee from an interest rate. And treat any national statement about "the housing market" with caution: housing is intensely local, and national aggregates routinely conceal simultaneous booms and slumps in different regions.

For the interest-rate background, see the economy; for the public-accounting side, see taxes and public finance.