I've done the math, and it adds up. Photographer: Chip Somodevilla/Getty Images
I've done the math, and it adds up. Photographer: Chip Somodevilla/Getty Images

Timothy Geithner claims in his new book, "Stress Test," a memoir of his years as U.S. Treasury Secretary during the financial crisis, that the federal government has turned a profit on its bailout of the banking system.

It's an assertion offered as one of the many reasons the rescue was a good idea. Let's not debate whether the bailout was or wasn't necessary: Get real -- it was. But Geithner's assertion about the bailout's cost, or rather lack thereof, demands scrutiny.

Specifically, Geithner says the $700 billion bailout has produced a $51.8 billion gain for the government. About half of that, he says, came from the rescues of the mortgage-finance companies Fannie Mae and Freddie Mac, which now are profitable and paying dividends to the feds.

Geithner's claim seems to rest on a straight-up accounting, much like this one at ProPublica, which to date shows a 5 percent profit on the bailout. Never mind that this is a return on taxpayer money for the past 5 1/2 years, a trifling amount really. Nor does it reflect what else could have been done with the money.

There are other problems with Geithner's claim, among them that it doesn't square with the White House's own numbers. As of Sept. 30, the end of the latest fiscal year, the Troubled Asset Relief Program was a loser for the government, according to the Office of Management and Budget, even if the size of the loss was headed in the right direction:

The direct impact of TARP on the deficit, including interest on reestimates, and the risk-adjustment to the discount required under (the bailout), is projected to be $47.5 billion, down $20.5 billion from $68 billion as projected in the 2013 (midyear review). The subsidy cost represents the lifetime net present value cost of TARP obligations from the date of disbursement.

In other words, Barack Obama's administration expects to lose money on the bailouts. The Congressional Budget Office, using a slightly different approach, reached the same conclusion, estimating the loss to taxpayers at $21 billion.

But these are all measurements that fail to capture perhaps the biggest cost: the transfer of capital from savers to lenders.

One way to look at it, as noted by Edward Harrison at Global Macro Advisors LLC, is based on personal interest income as tallied by the Federal Reserve Bank of St. Louis and illustrated in the following chart:

Until the financial crisis erupted in September 2008, personal income had climbed, albeit with some ups and down. As the Federal Reserve cut overnight interest rates from 2 percent in mid-2008 to near-zero by the end of that year, interest earned by savers tumbled from an annual rate of almost $1.4 trillion to less than $1.2 trillion by September 2010.

Meanwhile, long-term interest rates fell too, but not by as much -- the 10-year Treasury has averaged about 2.7 percent since September 2008. In other words, the yield curve, which reflects interest rates for different maturities, was steeply upward sloping, reflecting the Fed's monetary stimulus. Banks could borrow short term from depositors and pay next to nothing while lending for longer maturities at higher rates (or buying risk-free Treasuries).

This is the time-tested way of aiding banks at the expense of savers.

The data underlying this chart suggests that this transfer to lenders amounted to more than $150 billion a year on average. Multiply that by the 5 1/2 years since the onset of the crisis and you get a loss of $825 billion, with a lot of that going to the financial industry.

This figure isn't far off from another calculation of the subsidy by Richard Barring at MoneyRates.com. As he sees it, bank depositors traditionally earned interest that was a bit higher than the inflation rate. However, with rates near zero, depositors had an actual loss because of inflation. Here's his analysis:

A year ago, there was $9.427 trillion on deposit at U.S. banks. Over the past year, average money market rates have ranged from 0.08 percent to 0.10 percent. Inflation, meanwhile, was 1.5 percent over that same period. Because inflation grew faster than the average bank rate, consumers lost purchasing power. Adjusting that $9.427 trillion upward for interest earnings but then downward to account for the inflation rate yields a net loss in purchasing power of $122.5 billion. When this loss is added to the purchasing power losses from the previous four years, the total comes to $757.9 billion -- the effective price of the Fed's low-rate policies.

This probably is the other side of the same coin shown in the personal interest income graph. The bottom line, though, is the same -- TARP was hardly the only subsidy for the banks.

Of course, there are all sorts of other costs associated with the financial crisis -- the economic growth that wasn't realized, the erosion in human capital from long-term unemployment, the evaporation of household wealth. And no one should deny the incalculable costs that were avoided by the actions Geithner took to head off a full-scale economic collapse.

But justifying the bailouts because a simplistic measure shows the government made money is a stretch.

To contact the author of this article: James Greiff at jgreiff@bloomberg.net.

To contact the editor responsible for this article: Lisa Beyer at lbeyer3@bloomberg.net