3.4%, NOT 34 Basis Points

With thanks to jck at Alea, I retract part of what I wrote earlier, but leave it up for all to see my mistake.  3.4% is correct for the ROA.  Now, note if you are levered 44 times, 3.4% means you earned ~50% more than your capital in a year, which is quite a haul.

Post-embarrassment for me, 3.4% is reasonable, and more in line with what I thought my topic would be about when I started writing last night.  The Fed is taking risks:

  • Duration — short liabilities, long assets.
  • Convexity — mortgages are prepaying
  • Credit/Equity — GSEs (modest), Maiden Lanes, AIG

Will they prove to be prudent risks? Not sure.  The duration risk concerns me the most.  Now the Fed doesn’t have to mark to market like the rest of us mortals, and having 3.4% of ROA covers a multitude of errors.  It would engender a loss of confidence if the yield curve got really steep, that the Fed was insolvent on a mark-to-market basis — that could lead to a loss of confidence in the dollar.

But that would take more than a 2% move up from here, and who knows?  Maybe the Fed would suck in a lot of high coupon, long duration treasuries at that level, and profit off of a bigger spread, kind of like the attitude some life insurers had in the mid-80s.  Mismatch long to earn extra money, and if rates rose higher, write more business to get out of distress caused by unrealized capital losses.  Eventually won’t work if you have risk-based capital rules, or liabilities that can run, but if you are the Fed, then there are almost no limits except that of inflation kicking in amid economic weakness.

Following this is the piece with the math error, as published.


Who cares what the Federal Reserve remits to the US Treasury?  Some are amazed by the record $78.4 Billion remitted to the US Treasury, as if that were “found money.”  It is all stolen out of the pockets of savers, who deserve a currency that is truly a store of value, and rewards saving.

The amazing thing to me is that the Fed only earns 34 basis points on assets, especially since they pay less than 25 basis points on assets as a cost of funds over the last year.  The huge earnings points me to the massive leverage the Fed uses, 44x, which it would/should not let the banks it regulates operate at.

But the Fed is different, because it can create money/credit out of thin air.  It can’t go broke as a result.  But, in the limit, that doesn’t mean that really ugly things could not happen if the Fed were to create money/credit with abandon in order to make its books balance after massive capital losses.

So I don’t give the Fed any credit for remitting a record amount to the US Treasury.  The day may come when the US Treasury may have to recapitalize the Fed after credit losses, or the day may come when inflation causes people to distrust the value of the US Dollar.

My big surprise is that the Fed isn’t earning more in this environment.  QE 1&2 should be rich sources of earnings, but is it not happening because the Fed keeps its maturities short?  If so, good, and it explains why QE is so weak.  QE is meant to stimulate through lowering longer interest rates, and that has not happened to the degree that it might, which means the Fed is playing it safe.

Humph. So we have a Fed that muddles in the middle.  Probably the best that we could hope for?  Perhaps.  I need to think about this more, and perhaps this mutes my criticism of the Fed.  That’s not what I intended when I started writing tonight, but it is where I am now.


  • matt says:

    Can you clarify something about Fed operations for me. You said, “…they pay less than 25 basis points on assets as a cost of funds over the last year.”

    My impression is that when assets are added to the Fed’s balance sheet, dollars are simply “created” and credited to the sellers’ accounts. These dollars are notes that effectively cost them 0. How is there any cost of funds for the Fed when it makes asset purchases?

    Also, my impression is that the Fed is standing on the front end of the curve as part of the stealth bailout of banks by keeping the yield curve steep. I suspect Fed Open Market Operations would, in aggregate, look very differently sans this motive.

  • I believe the Fed is presently paying between 0.00% and 0.25% to have banks hold their reserves at the Fed at present.

  • Greg says:

    The Fed’s balance sheet is more than twice the size of last year, but their “remitances” to the US Treasury are up only 90%?

    If they had a constant profit margin on assets, one would expect remitances to also be a bit more than double the previous year. Instead they are quite a bit less.

    Obviously there are some significant losses buried inside the claimed 3.4% ROA

    And if the Fed were being honest, many more losses were essentially moved to FNMA/FHLMC’s balance sheet instead of being recognized at the Fed.

    Anyone can get 3.4% ROA (probably even a bank CEO) if they count gains but shift losses to another entity

    Two Christmas Eve’s ago, Geithner made it very clear that FNMA / FHLMC is where they plan to hide all the bodies.

    Its pretty easy to fool the average member of the public with fancy accounting gimmicks. Its a bit more surprising that finance professionals can’t see through the scam (or they are choosing to be fooled). This tells us a lot more about the health of Wall Street than it does about the Fed’s profits

3 Trackbacks