Remarks at the COMMIT Forum, New York, NY. October 3, 2012
Before talking about either the impact or the need for Sarbanes-Oxley, I think it would be helpful to briefly review what the Act covers.
Despite all the talk about transparency and disclosure, most of Sarbanes-Oxley is about the regulation of auditors, from the creation of the Public Company Accounting Oversight Board to requirements on auditor rotation and restrictions on the provision of non-audit services.
Yes, there is title 4, the one out of 11 titles, that deals directly with disclosure and perhaps the most infamous section, 404, which requires management to attest to effectiveness of internal controls.
There are also increased criminal penalties for fraud and other white collar crimes.And of course, it wouldn't be Congress if there weren't several studies and reports.
I will note here, that despite the failure of the credit rating agencies in detecting problems at companies such as Enron and WorldCom, highly rating their debt until just before their collapse, and the role of the rating agencies in the recent mortgage crisis, the approach of Sarbanes-Oxley was requiring a study, rather than addressing the fundamentally flawed regulatory structure, which had turned the market for ratings into essentially a duopoly.
This is a theme to which I will return: the choice was not Sarbanes-Oxley or do nothing, just as our recent choice was not Dodd-Frank or do nothing. Our choice could have been to actually address the flaws in our financial and monetary system, which I believe we did not. The heart of SOX is subjecting auditors to both a registration requirement and to examination by the PCAOB. It is worth remembering that a similar scheme for investment advisers did not prevent the Madoff or Stanford schemes.
The predictable outcome is that SOX's increased regulation of auditors would reduce competition and increase concentration among auditors. In fact, academic studies have found that post-SOX, for Fortune 1000 companies, their average audit fees increased 67%.
In my opinion this is probably the most bizarre aspect of SOX: you believe the auditors misbehaved, so you by law increase the demand for their services and erect substantial barriers to entry that massively benefit the same incumbent firms which you claim were behind the crisis.
Again let me emphasis that despite SOX's ban on auditors performing non-audit work for the same companies, audit fees actually went up. Even with a decline in non-audit business, the audit fees charged increased more than enough to off-set that decline. And not surprising, given SOX's reduction in competition across auditors, auditing fees went up the most among the largest firms. But then a basic understanding of economics would have predicted such.
Now this increased cost might be worthwhile if audit quality increased. If anything the opposite appears to be the case. Reported instances of corporate fraud have increased post-SOX.
The reports of the PCAOB suggest that SOX itself may have contributed to a decrease in audit quality. One of obsession of SOX is increasing auditor independence by forcing a company change auditors regularly. However deficiencies reports from PCAOB indicate that a longer tenure for the audit reduces deficiencies, rather than increasing them.
Auditor independence is not free. It comes at the price of auditor understanding of both the company and industry in question. By deciding in favor of a contrived version of independence, SOX has embraced auditor ignorance.
It is worth noting that the issue of auditor independence had been subjected to repeated analysis in the academic literature. The conclusions of that literature so contradict the provisions of SOX that Yale Law Professor Roberta Romano labeled them as "quack corporate governance".
Let me touch for just a moment upon SOX's infamous Section 404, which requires management to attest to effectiveness of a company's internal controls, and to have those internal controls evaluated. First does anyone really doubt that a CEO intending to commit fraud would be deterred? If 404 had been in place, I think it's a safe bet that Ken Lay, CEO of Enron, or even Frank Raines, CEO of Fannie Mae, would have happily attested to their firms' internal controls.
Now in the case of Fannie Mae, where was also considerable fraud, it wouldn't have mattered as much, because despite his professed commitment to disclosure and transparency, Senator Sarbanes repeatedly protected Fannie Mae from having to comply with the very securities laws he believed other companies should be subjected to. If one believed that such laws were for the benefit of investors, it is hard to believe that investors in Fannie Mae didn't merit the same protections, not to mention protecting the taxpayer.
Which brings me to one of the fundamental flaws of Sarbanes-Oxley and more generally, our body of securities laws: usually it isn't management that pays, it is the shareholders. Or to be more exact, the plaintiff lawyers bring a shareholder class action, management agrees to have the firm pay previous shareholders, at the expense of current shareholders and of course, lawyers take a huge cut off the top.
This is the nature of SOX. It creates impossible, subjective standards that can always be contested after the fact, so that enterprising lawyers can make off like bandits while investors get left holding the bag.
Of course we are told the opposite that investors are better off under SOX. I will come back to whether such a claim withstands empirical testing, but if indeed investors are made better off then why wouldn't firms voluntarily adopt those corporate governance reforms that increase shareholder value, as doing so would make it easier to raise capital. In many cases, they do.
We should remember that it wasn't the SEC that created financial disclosure or good corporate governance. It was a combination of demands by then-private exchanges and corporations seeking charters that were attractive to investors.
Before SOX and even before the creation of the SEC, corporation law was largely the domain of the states. This competition among the states allowed for a process of trial and error, as to what works best to protect investors. If states had poor corporate governance regimes, such was reflected in a depressed share price.
SOX takes the arrogant approach that its authors know best and imposes a one-size-fits all regime on all US corporations. A little more modesty would be in order, as no one knows ahead of time the optimal corporate governance framework. To truly protect investors, the provisions of SOX should allow corporations to opt-out by a shareholder vote. Such would demonstrate whether it was valued or not.
As with most of what comes out of Washington, defenders of Sarbanes-Oxley regal us with its many imagined benefits, but conveniently ignore its costs. But there are no free lunches in public policy. Any piece of legislation has both costs and benefits; it is the net benefit that matters.
From the SEC's own estimates, publicly traded firms spend on average over $2 million annually in direct costs compiling with Sox. Estimates from economists at the University of Rochester conclude that Sarbanes-Oxley reduced shareholder wealth by about $1.4 trillion. This was supposed to be about helping investors.
SOX has also had the direct role of reducing the competitiveness of America's capital markets. Companies around the World used to flock to US exchanges. There has historically been a significant benefit for foreign firms to cross-list stock on a US exchange. After SOX that premium significantly declined.
This impact shows up not only in stock values but in the willingness of firms to go public on US exchanges. Prior to SOX, US IPOs averaged 27% of global IPOS, since SOX that percentage has fallen to 12%.
SOX has not only reduced the number of firms going public, it has also encouraged firms "to go dark" that is de-list their shares. The number of such voluntary de-listing more than doubled after the passage of SOX. If the proponents of SOX had hoped to protect investors with more mandated disclosure, the result has likely been a lower level of disclosure, on average, as more firms choose to either remain private or to de-list.
Perhaps the most damning fact about SOX is not its costs and its negative impact on capital formation, but the simple and indisputable fact that it neither eliminated corporate fraud nor ended financial bubbles and panics.I mentioned earlier the SOX approach to credit rating agencies was to demand a study. It is also worth remembering that what sunk Enron was a variety of off-balance sheet vehicles that obscured the real risk Enron was taking. But there's fundamentally little difference between these accounting games and the off-balance sheet vehicles used by say Citibank, post-SOX, to warehouse sub-prime mortgages.
The most fundamental flaw behind SOX is its absolute avoidance of addressing asset bubbles. Let us recall the problems at Enron, WorldCom and others did not come to light until after the bursting of the dot-com bubble. These discoveries of fraud did not cause that bubble to burst. What did cause that bubble to burst was a reversal in the flood of easy money by the Federal Reserve that drove up equity prices.
Despite his concerns about "irrational exuberance" Greenspan pushed the monetary supply to increase at a rate far in excess of the growth of the underlying economy. Out of misplaced fears from a Y2K problem the Fed flooded the financial markets with liquidity. That money went into the stock-market. When the Fed began to reduce that liquidity, the bubble burst. Again the failure of Sarbanes-Oxley is to believe that isolated instances of fraud were the driver of a boom and bust, rather than poorly managed macroeconomic and credit policy.
SOX's attempt to restore investor confidence, without actually changing the fundamentals of our financial system, only leaves investors all the more vulnerable next time.
Lastly we should remember that the actions which Sarbanes-Oxley was a reaction to, the financial frauds at Enron, WorldCom, and others, were already crimes before Sarbanes-Oxley. In fact several individuals were convicted and sentenced under laws that existed before SOX. At the time of his death, Ken Lay was facing a sentence of up to 45 years.
Putting more crimes on the books simply because other crimes were committed, hardly seems like an effective deterrent.
Finally it is worth noting that SOX's mandated changes to corporate boards were already standard practice among US corporations. In fact, Enron's board, before its failure was already compliant with SOX. It would appear that proponents of SOX believe corporate boards should be more like Enron's, rather than less.