Fannie Mae: the implied story, Part 3: Gaming the bonus formula
[For the introduction, see Part 1.]
[For previous installments, see Part 2.]
[Quotes in green are from the OFHEO report (full document, .pdf), on which Fannie Mae declined to comment. Sentences are occasionally reformatted (bullets, boldface emphasis, inconsequential brief elisions) for clarity.]
So far we have had the overture (Part 1), and introduced CEO Franklin Raines, whom OFHEO is auditioning for villain (Part 2). Now, what was the game?
Quick, Watson, the game’s afoot!
To maximize core business earnings per share (hereinafter EPS):
Under the Fannie Mae executive compensation program, senior management reaped financial rewards when the
Performance was measured by a target set by the players, measured by the players, and refereed by the players. Not bad.
I’ve never failed since I started cheating at solitaire!
Some of us who have spent our entire careers in finance would wonder how EPS could rise so smoothly. After all, when interest rates rise and fall, the value of financial instruments rises and falls even faster. Had Fannie Mae found the perfect gimbals to keep their earnings smooth?
Next you’ll be finding the longitude.
In finance, we are not supposed to set our own scorekeeping rules: accountants do that, under standards established by the Financial Accounting Standards Board (FASB, rhymes with razz-me) that as required updates the Generally Accepted Accounting Principles (GAAP). Fannie Mae found GAAP too constricting, so when inconvenient new Financial Accounting Standards (FAS) issues came out, Fannie Mae elected to ignore them:
Ah find financial accounting standards so … restricting!
The period-to-period volatility in reported net income that resulted from the requirements of FAS 133 did not, in the view of
When the FASB told the marketplace to use FAS 133, Accounting for Derivative Instruments and Hedging Activities, Fannie Mae told its board and investors that this would apply inaccurately to Fannie Mae, and got approval to use its own measure. (By the way, even though Fannie Mae’s shares are publicly traded, until recently the company was not required to list with the SEC, so the SEC had no means of compelling GAAP compliance. This has now changed.)
Le market, c’est moi.
Ignoring FAS 133 was just one of several exceptions Fannie Mae sought for itself. It skipped over FAS 91 (Accounting for Nonrefundable Fees and Costs Associated with Originating or Acquiring Loans and Initial Direct Costs of Leases), entered into income-shifting transactions (more on these in Part 5), and in general worked backwards from a desired EPS goal:
By utilizing the strategies described above as a foundation, Fannie Mae management was in a position to employ several techniques to manipulate and manage earnings more directly. Those strategies included the use of cookie-jar reserves, certain Real Estate Mortgage Investment Conduit (REMIC) transactions to delay federal taxes or defer earning recognition, debt repurchase, and certain insurance transactions. Page 7.
Those reserves and transactions were utilized and maintained to provide management with the opportunity to make last minute quarter-end adjustments to hit specific earnings targets. Page 7.
Just in case you’ve forgotten why the focus on earnings, OFHEO reminds us:
The transactions and strategies constituted additional instances of inappropriate earnings management undertaken to achieve annual EPS targets and maximize bonus payouts to senior management, violating safety and soundness standards. Page 7.
One might think that all this emphasis on earnings, however aggressive, would result in them not just hitting targets but exceeding them — but earnings were also managed down as well as up:
Furthermore, the Annual Incentive Plan provided no incentive for management to add to earnings once the EPS number for a maximum bonus payout was achieved. That encouraged the shifting of income forward in years of plentiful core business earnings to meet EPS targets in future years as well. Page 5.
Year after year, even quarter after quarter, the targets were hit with extraordinary accuracy:
In the following years, time and time again Fannie Mae employed adjustments that enabled it to meet its EPS targets, whether on a quarterly basis to meet analysts’ expectations or on an annual basis to meet compensation targets. Page 41.
“We’re good at this.”
All right, says the by-now-inundated reader, I concede that gaming was taking place. Aren’t Earnings Per Share a good thing to measure? Isn’t that what stockholders want?
Executive compensation expert David Yermack, a
EPS is very much in disrepute among academics (as well as compensation professionals) because it is a statistic that managers can manipulate very easily. Research as far back as the 1980s has shown that managers ‘smooth earnings’ across time quite aggressively, and it has never been done more obviously than in the situation involving Freddie Mac. The managers at Fanne [sic] Mae, and just about everywhere else, should have their income compensation linked to the stock price, not to accounting targets. Page 69.
Moreover, Earnings Per Share equal total earnings divided by total shares outstanding. So let’s just say we embark on an aggressive debt or stock buyback campaign, using corporate cash.
“Let’s just say … we’re gonna use this!”
That is, we take our cash, and we buy our own securities; in doing so, we can create earnings or losses:
- If we buy debt for more than par, we take a loss.
- If we buy it for less, we book earnings.
Meanwhile, if we buy shares, we lower the denominator and increase the Earnings Per Share.
Just a little routine financial engineering, ma’am.
There is no question Fannie Mae used both levers:
While debt buybacks offered powerful opportunities for adjusting earnings, stock buybacks were very effective for fine-tuning EPS. In notes prepared for a November 1996 board meeting, Mr. Howard explained how EPS, which by then had a direct impact on the onus amounts paid, could grow at a substantially greater rate than net income:
You’ll see that for the four year period we’re projecting average EPS growth of nearly 12 percent—11.8% to be exact. Net income growth is a bit under 10%. The difference between the 9.8% net income growth and the 11.8% EPS growth is the assumed effect of the continuation of our stock buyback program. Page 44.
This is the same Mr. Howard who told Congress, under oath, that incentive compensation played no role in Fannie Mae senior executives’ decision-making.
In another exchange Baker cited, Rep. Barney Frank (D-Mass.) asked Raines and Howard, “You both deny that trying to hit a certain amount so you could get your bonuses was a factor to any extent in your decisions?”
“We both deny that,” Raines replied.
“Yes,” Howard said.
OFHEO’s report repeatedly contradicts these unequivocal statements:
Mr. Howard was not the only executive at Fannie Mae who understood the correlation between the stock buybacks and EPS. In the 1997 Performance Assessment for Mr. Howard, COO Lawrence Small encouraged Mr. Howard to “speed up the pace” of the stock buyback program, but added a very specific parameter for the “pace”:
Obviously we recognize that our buyback pace has to be calibrated to fit our desired EPS growth rate, so don’t take the previous statement as anything more than strong encouragement to stay focused on this important aspect of capital management. Page 44.
You can’t see what we’re doing, can you?
This is unbelievably transparent stuff, but if you think that is transparent, try this:
“You’ll see that for the four year period we’re projecting average EPS growth of nearly 12 percent—11.8% to be exact. Net income growth is a bit under 10%. The difference between the 9.8% net income growth and the 11.8% EPS growth is the assumed effect of the continuation of our stock buyback program.” Page 81.
More than 15% of the EPS boost (11.8% / 9.8%) is from stock buyback.
“In the forecast we assume that we repurchase the six percent of outstanding shares authorized in our December 1996 capital restructuring program evenly over 1997, 98 and 99. This has two effects. First, the additional debt costs we incur to repurchase the shares reduce our net interest income, and also our net income, by about one percent. (That is, were it not for the buyback, our net income would grow by about 11%, rather than 10%.) But the effect of having significantly fewer shares pushes our EPS growth above the 11 percent growth we would have had without the buyback, to the 11.8% shown here.” Pages 81-82.
As described, this is a transparent redistribution from shareholders to executives:
- Shareholders own the cash used to buy back the shares.
- Executives receive bonuses based on a change in the Earnings Per Share metric, rather than an actual change in value of the company.
Notice how I never take my hands off the cash!
Inferring from the record, all heads — themselves likely beneficiaries of one or more of the three rings of substantial bonus executive compensation — nodded approvingly at Mr. Howard’s plan. Evidently this kind of thinking was pervasive:
[Controller Leanne] Spencer also saw the connection between stock buybacks and EPS. In her Quarterly Business Review talking points for the first quarter of 2000, under the heading “Double Income Goal,” Ms. Spencer referred to a chart that summarized what each business segment would be assigned to achieve the goal of doubling EPS by 2003. She noted, “[f]ortunately, each business segment doesn’t need to quite double in order for us to meet our $6.46 goal, because of the benefit of stock repurchases and other corporate actions. We are calling that amount our ‘contingency’ reserve.” Page 44.
Clearly, Ms. Spencer and other Fannie Mae senior executives viewed stock repurchases as a type of contingency reserve to enable the
Returning to OFHEO, stating what you will now see as obvious:
While companies typically link the compensation of their executives to firm performance, relying heavily on one accounting-based measure such as earnings per share is problematic. Academic literature and practical experience suggests that when such a linkage exists executives can and do act aggressively to maximize their compensation by making accounting adjustments. Page 5.
Theory met reality, early and often:
If Fannie Mae’s earnings exceeded amounts necessary to hit EPS targets, senior management cast about for transactions of marginal business purpose that had the effect of moving income from the immediate period, where it was not needed, to a future period, when it might be needed to hit EPS targets and maximize bonuses. Indeed, as discussed in Chapter V, excess earnings per share above those targets did not result in additional bonuses. Page 43.
One other feature of the metric deserves mention: it wasn’t all Earnings Per Share, but core business Earnings Per Share, so some earnings and losses could be left out. Guess which ones got used at the margin?
The economic losses associated with those [portfolio exposure, see Part 4 -- Ed.] rebalancing actions were not reflected in Fannie Mae’s core business earnings in the periods in which they were realized. Page 47.
For example, prior to registering with the Securities and Exchange Commission, the
In other words, the vast bonuses paid or accrued, which themselves are reductions in EPS, are not counted as EPS. How handy!
Fannie Mae’s core business EPS was the One Ring of Compensation, saying in financial script:
One means of keeping score
One way of paying
One type of earnings game
Taxpayer’s risk off-laying
Am I vested in this option?
Now, how to maximize those earnings?
[Continued tomorrow in Part 4.]
