I don't have much insight into Citi's earnings but I do enjoy the reporting of them. When a car or Facebook company reports earnings you basically ask questions like "how many cars or Facebooks did it sell?" and "how much money did it make on each one?" and those questions are kind of answerable and their answers give you a sense of how you should feel in your heart about the company. When a bank - like, a bank bank - reports earnings you can ask "how many mortgages did it sell?" and "how much money did it make on each one?" and those answers will be useful to you too, though there will be murky liquidity and valuation overhangs that will reduce their usefulness.
If you asked those questions of Citi, you might or might not get answers that might or might not be useful, but you'd be hard pressed to translate them into the headlines on Citi's earnings. Big banks are not primarily engines for selling products and collecting a margin on them; they are bundles of accounting decisions, and this is never more apparent than at earnings time. This is pretty far removed from economic activity in the world:
Citigroup Inc.'s third-quarter profit fell 88% as the bank took charges tied to the value of its debt and the sale of a stake in its brokerage joint venture ...
Others chose to emphasize economic activity in the world, at the cost of, y'know, GAAP:
Citigroup Profit Beats Estimate on Bond-Trading Gains
Take your pick! It is fun to say things like "non-cash accounting charges are fake and you should back them out and concentrate on recurring items with economic significance" but that's just, like, your opinion man. I can't really defend DVA as a thing to care about - I mean, I sort of have, but that was mostly out of perversity - but the charges on Morgan Stanley Smith Barney are totally, totally, totally real! They're not really this quarter - Citi didn't like lose $4.7 billion of actual money on Smith Barney in July through September 2012 - but they're real: Citi bopped around for a long while saying "oh yeah we've totally got a $23 billion retail financial advisory business" and turns out they didn't, so, yeah, sure, somebody misplaced some money somewhere, though of course the writedown is not a surprise now.
If you wanted a model of how to react to Citi's earnings you could I suppose build one based on your subtle and comprehensive understanding of real economic factors. Things like actually growing sustainable lending businesses would look good, things like sharp surprising improvement in FICC trading would cause skepticism in the short term until you fully satisfied yourself that this is "Citi is expanding its FICC client franchise and/or getting smarter at trading bonds" rather than, like, the coin came up heads this quarter. You'd build your own model for MSSB valuation, based on its performance, and Citi's accounting value for MSSB would carry no weight in your price. That accounting value of MSSB, like the noise of DVA, would just be meaningless presentation issue, no more important to you than the font that Citi uses to report earnings.
That would be great, but also, like, hard. Here's a simpler idea. This is Goldman's quick research take this morning:
Basel 3 capital increased to 8.6% from 7.9% last quarter on capital accretion from the MSSB sale, continued runoff in Citi Holdings and retained earnings. Encouragingly, Citi is now within range of it’s fully phased in Basel requirement after factoring in additional capital accretion to come from the remaining MSSB sale (~+40bp). ...
We expect shares to respond favorably today given strong top-line performance, positive operating leverage, and better than expected Basel III capital. We expect the call to focus on (1) the outlook for capital return (2) the pace of future asset run-off in Citi Holdings and (3) outlook for 4Q & beyond in the securities business. We maintain our Buy rating and make no changes to our price target.
The "shares to respond favorably" prediction was right, but I particularly appreciated the focus on capital and capital return. Basel III capital is, like, an even faker thing than GAAP earnings, isn't it?1 But unlike GAAP, it is readily translatable into actual economic activity in the world. Citi, you may recall, wanted to give out money to shareholders, but was stymied by its regulators' view that doing so would leave it without enough capital to deal with stress. Now that it has more capital than expected, it has a better chance of giving money back to shareholders.
That would be good. As of September 30 Citi's equity is worth $179.3 billion as far as US GAAP is concerned; as of this sentence it's worth $106.6 billion as far as the stock market is concerned. A dollar in Citi's hands is worth 59 cents: the sooner Citi gives you back your dollars, the happier you'll be.
Citi earnings 8-K [EDGAR]
Citigroup Profit Beats Estimate on Bond-Trading Gains [Bloomberg]
Citi Profit Slides, but Core Businesses Show Strength [WSJ]
Citigroup Earnings Plummet in Third Quarter [DealBook]
1."No" is a fine answer there. Basel III is a collection of accounting conventions, and so is GAAP, and that's pretty much that. Each has pluses and minuses. The fact that you can make up a lot of your Basel III risk weighting might worry you, but on the other hand Basel III, for instance, excludes DVA.