“Regulating” and “deregulating” are terms that often mislead. My next three columns discuss how to regulate two diverse activities that are critical to our economy – residential mortgage lending and starting small businesses. This column explains the most regulatory approaches essential to regulate residential mortgage lending effectively. Next week's column will discuss why the regulatory approach we have taken and the modifications to that approach contained in Basel III do not provide an inherently unsound regulatory structure. The third column will deal with regulatory structures that aid small business formation.

Effective regulation must begin with the rationales for regulating the activity. The failure to take this approach was critical to the crisis we have just experienced. One of the great failures was assuming that if the lender was not federally insured there was no need for federal regulation. That assumption, in turn, was based on assumptions about the type of institution requiring deposit insurance. Both of those assumptions are large topics with voluminous literatures that will be the subject of future columns. For purposes of this column I assume that we have decided that the federal government should regulate residential mortgage lenders. Most of the principles I discuss also apply to commercial real estate (

) lending (which includes loans to build more than four residential units), but CRE has some unique characteristics that warrant a separate column. This column focuses on safety and soundness regulation as opposed to compliance, but I emphasize that effective enforcement of rules to protect borrowers would have prevented trillions in dollars in losses to lenders. Indeed, that example exemplifies my central point – effective regulation is essential and desirable to protect honest lenders. That does not mean that all regulation is desirable or that more regulation is better than less regulation. Our central function as financial regulators is to reduce criminogenic environments and prevent epidemics of accounting control fraud. Home mortgage lending is an industry that we know how to do well. Historically, credit losses on home loans – from all sources – have been under one percent. That means that residential mortgage lenders have long understood how to limit fraud losses to well under one percent. The good news is that the same rules that dramatically limit losses from imprudent loans are exceptionally effective in preventing fraud. U.S. home lenders suffer severe losses in three circumstances: due to sharp, sustained increases in interest rates, accounting control fraud, or the collapse of hyper-inflated residential real estate bubbles. Foreign banks can also suffer severe losses due to currency risk. U.S. mortgage loans are made in U.S. dollars and borrowers' salaries are overwhelmingly paid in dollars, so this column does not address how regulators should respond to currency risk.

Home lenders can also fail due to poor cost controls relative to their competitors, but these failures do not cause serious losses and do not pose systemic risks. These failures also typically require several years to occur and are simple for the regulators to spot through routine reviews of the banks' “call reports.” We send examiners in to confirm the reasons the lender's general and administrative expenses make it uncompetitive, but the problem is almost always weak managerial skills. We try to convince the bank to hire new managers or find an acquirer before the failure. Our great advantage as regulators over other entities that are supposed to correct such problems, e.g., the board of directors or the outside auditor, is that we can be truly independent. The board of directors was picked by the CEO and signed off on the business strategy that is leading the bank toward failure. The audit partner fears that he will lose the client if he gives a negative audit opinion. It is not the auditor's function to serve as a business consultant. Good regulators can help in this sphere, but this is not the sphere in which we must show great courage and it is not the sphere in which we can prevent hundreds of billions of dollars in losses, Great Recessions, and the loss of over 10 million jobs.

The important exception to this conclusion is that courageous regulatory intervention is essential where the banks' failures to be competitive is caused by market power and implicit federal subsidies to banks deemed “too big to fail.” Systemically dangerous institutions (SDIs) are far less efficient than their competitors, but they can obtain decisive advantages over smaller competitors because of their ability to borrow more cheaply. That competitive advantage arises in some regions from their market power, which allows them to raise deposits at lower interest rates. Because they are perceived as “too big to fail” they, SDIs are the beneficiaries of an implicit federal subsidy that allows them to borrow other funds more cheaply than smaller competitors. I'll deal with the necessary regulatory steps to rid us of the SDIs, which are inefficient and dangerous, in future columns. The Bush and Obama administration policies toward the SDIs have made them far larger, substantially increased their market power, and increased the systemic risks they pose.

Interest rate risk can also pose systemic risks. There is no reason for a home mortgage lender to take large interest rate risks in the modern era. They can transfer the risk either by selling the home loans (and hedging the pipeline) or keeping the home loans in portfolio and hedging the risk. There is no societal advantage to mortgage lenders taking substantial interest rate risk (the expected value of taking interest rate risk should be zero). The Special Investment Vehicles (SIVs) that the huge investment and commercial banks created to hide their sister banks' true debts added exceptional interest rate risk to their overwhelming operational (control fraud) risk. The Regulators, therefore, should not allow lenders or their affiliates to take substantial interest rate risk and should not allow bank holding companies to create SIVs. SIVs create substantial systemic risk and make finance opaque. From society's standpoint, SIVs are unambiguously harmful. Identifying, measuring, and controlling the interest rate risk of a portfolio of American mortgages is a particularly complex process because of the embedded prepayment option and the lack of prepayment penalties. Hedge accounting is notorious for its abuses. The SEC charged that Fannie's controlling officer abused hedge accounting to inflate reported earnings to maximize their bonuses and that Freddie's controlling officer manipulated hedge accounting to create “cookie jar” reserves that they could draw on whenever desirable to maximize their future bonuses. Many purported “hedges” are actually designed to prevent loss recognition and involve speculative investments that increase interest rate risk. The banking regulators and the SEC can improve their chances of detecting these scams by having the examiners (with appropriate accounting support) check to ensure that the lender is keeping contemporaneous records documenting that the purported hedging instrument was actually purchased to hedge a specific position and that the bank had demonstrated and documented that the instrument would function as an effective hedge. This discussion illustrates one of the essential facts about effective financial regulation – enforcing honest accounting is a prerequisite. “Dynamic hedging” is an oxymoron that is subject to even worse abuses than conventional hedges. Dynamic hedging cannot protect against large changes in interest rates (which are the changes we most need to worry about as regulators) and can cause a severe systemic risk that can drive market crashes. Regulators, therefore, should prohibit “dynamic hedging.” The good news about regulating interest rate risk is that if the regulators do ban dynamic hedging, ensure accurate records of real hedges, and forbid banks from taking substantial interest rate risk then it is very difficult for a bank to take large gambles on interest rates. There is no societal benefit to banks taking substantial interest rate risk in the modern era. An honest, competent bank would not take serious interest rate risk and purport to use dynamic hedging to neutralize that risk. An honest, competent bank would test and document its hedges. Honest, competent banks would do all these things even if there were no regulators. Bankers, not regulators, devised these business practices because they are essential to running a prudent bank that can prosper and survive. This discussion of the procedures that competent banks would follow in the absence of banking regulation also illustrates why studies purporting to show immense compliance costs to banking regulators are false. Most of these costs falsely classified as costs of regulation would be borne by banks regardless of whether the regulator existed. Other costs, such as creating the call reports, are costs of regulation, but they are of great value to the banks. Absent the regulatory requirement to provide the data and the role of government examiners and data specialists in keeping the data more comprehensive, comparable, and accurate the banks would have to pay a private sector entity to create an inferior industry data system, likely at greater cost. Banks that are exposed only to modest interest rate risk take a long time to fail even if interest rates increase sharply. America, relative to other nations, has had low interest rate volatility. This means that American banking regulators have typically had ample forewarning of problems arising from interest rate risk. Losses due to interest rate changes have not driven modern American bank failures.

Epidemics of accounting control fraud have driven our two most recent U.S. financial crises (the S&L debacle and the current crisis as well as the Enron era frauds). The national commission that investigated the causes of the S&L debacle reported that at “the typical large failure” “fraud” was “invariably present.” (The S&L debacle was a tragedy in two acts. The first act was driven by interest rate risk and it caused serious losses. The second act, which proved roughly five times more expensive than ultimate losses from interest rate volatility, was driven by the accounting control frauds.) The current U.S. crisis was driven by far more extensive mortgage fraud led by the large nonprime specialty lenders. The incidence of fraud was so great that it hyper-inflated the largest financial bubble in history.

Accounting control fraud is so dangerous because it simultaneously attacks the greatest weaknesses of the private and public sectors. To see why we have to reprise the four-part recipe for lenders maximizing (fictional) short-term income: 1. Grow extremely rapidly 2. By making high yield loans to those who will often be unable to repay 3. While employing extreme leverage 4. And providing only grossly inadequate loss reserves (ALLL) As Akerlof & Romer stressed in their 1993 article, accounting control fraud is a “sure thing.” They entitled their article – Looting: the Economic Underworld of Bankruptcy for Profit in order to emphasize that the same fraud scheme that produced huge (fictional) income maximized real losses and was a leading cause of bank failures. The fictional income also made exceptional compensation to the bank's officers a sure thing. The lender fails (in the era in which Akerlof & Romer wrote – we now often bail out the frauds and leave their managers in charge), but the controlling officers walk away wealthy. The recipe makes individual control frauds into wealth destroying monsters that cause extraordinarily large losses to banks. That alone makes preventing and closing rapidly accounting control frauds our top regulatory priority. Unfortunately, epidemics of accounting control fraud are not rare, and such epidemics create perverse dynamics that cause vastly greater losses. I discuss the risks of such epidemics in greater detail below. The central problem is that accounting control frauds look wonderful to the public sector and inexperienced and ideological anti-regulators. Theoclassical economists assured regulators that lenders and shareholders' “private market discipline” makes accounting control frauds impossible. In reality, many lenders and shareholders rush to lend to and invest in banks reporting record profits – and the accounting control fraud recipe guarantees record (albeit fictional) profits in the near-term. The result is that creditors and shareholders lend and invest the cash that funds the fraudulent banks' exceptional growth. Theoclassical economists also assured regulators that independent experts, particularly top tier audit firms, would never give favorable opinions to fraudulent corporations. Law students were taught in their “law and economics” classes that they could safely rely on the auditor's opinion. In reality, the CEOs leading the largest accounting control frauds routinely hire top tier audit firms and consistently receive clean opinions blessing their fraudulent financial statements. The art of accounting control fraud is to suborn – not defeat – the internal and external “controls” and turn them into the most valuable fraud allies. The frauds use the auditors, appraisers, and rating agencies' reputation and seeming expertise to assist them in deceiving their investors, lenders, and regulators. Indeed, the CEO uses the initial expert's opinion, e.g., the appraiser, to assist him in suborning the next expert in the chain, e.g., the auditor. Regulators that rely on reported income, net worth, and losses are worse than useless against accounting control frauds. Unfortunately, that has become the norm in the financial regulatory world and the basis for the entire Basel process. It is an approach that cannot succeed. Accounting entries are subject to massive manipulation. It is common for the banks reporting the greatest profits to be massively insolvent. Standard econometric studies, during the expansion phase of a bubble or in the presence of accounting control fraud, produce systematically biased results that support the worst possible regulatory policies that optimize the criminogenic environment that attracts and optimizes the frauds.

Fraud begets fraud. The CEOs who control the lenders engaged in accounting control fraud deliberately create the perverse incentives that generate other frauds to aid their looting. Consider four examples of “echo” epidemics of fraud that produced the current crisis: • They generate endemic internal and external frauds by employing “liar's” loans • They generate endemic internal and external frauds by compensating their loan officers based on loan volume rather than loan quality • They generate endemic fraud by independent experts by creating a “Gresham's” dynamic. For example, lenders engaged in accounting control fraud may refuse to use appraisers who refuse to inflate the market value of the house. The lenders engaged in control fraud leak to the appraiser the loan amount so that the appraiser will know how high the market value of the home must be inflated. • They created a network of fraudulent suppliers of fraudulent mortgage loans by creating the perverse incentives that made fraudulent loan brokers the eager suppliers of fraudulent mortgage applications. The CEOs controlling the fraudulent lenders frequently optimized this echo epidemic by employing liar's loans and degrading their underwriting process so that it would approve tens of thousands of fraudulent loans. Echo fraud epidemics occur because the private sector is so responsive to financial incentives – including perverse financial incentives. A Gresham's dynamic does not have to corrupt everyone to be fully effective in the contexts discussed above. The CEO of the fraudulent lender or fraudulent loan broker only needs to suborn a small percentage of the appraisers and loan brokers to implement the first two ingredients in the accounting control fraud recipe. Private market discipline is exceptionally effective – in funding control frauds and generating echo fraud epidemics. The CEOs that controlled the lenders that created these perverse incentives knew full well that they would create endemic echo frauds. Their creation of an intensely criminogenic environment is sufficient to cause the echo epidemics of fraud.