Vikram Pandit isn’t known for his deal-making. The Citigroup chief executive famously got beaten by a nose by Wells Fargo in the battle for Wachovia two months ago. However, Pandit has seemingly pulled off one fantastic deal: Citi’s own bailout.
Some details of the deal haven’t been disclosed – and some haven’t even been nailed down. So piecing together what is going on is a bit like solving a Rubik’s cube with some of the squares missing. But, from what is public, Pandit has shuffled off to Uncle Sam much of the risk in his $306bn bad bank for what looks like a low insurance premium.
One key number, which Pandit let slip in an interview on the Charlie Rose television show last week, is that the original value of this bundle of assets was between $340bn and $350bn. Given that they are now valued on Citi’s books at $306bn, the assets have so far suffered a haircut of about 11%. That looks pretty modest.
It is impossible to be sure whether this is adequate without knowing exactly what assets are in the portfolio. Yet history suggests that banks’ valuations of mortgage-related assets have been like the throws of a drunken darts player whose every flight hits the ceiling – consistently off target and far too high.
Citi will absorb the next $29bn of losses. But even after adding that, there is only a buffer of about 20% between the original value of the assets and the value to which they would have to fall before the state started to get hit.
What’s more, the government isn’t just insuring Citi against losses. The Federal Reserve is providing the bank with a backstop loan to fund the portfolio net of all these losses, if they happen. That funding line means Citi isn’t under pressure to dump assets at fire-sale prices. Without the loan, the bank could have been in the position of many other financial institutions such as Merrill Lynch and UBS which had to crystallise big losses when they were forced to sell assets quickly.
In return for giving Citi so many goodies, the government receives $7bn in preference shares plus some warrants to buy ordinary shares. That doesn’t look much. One pointer is the deal the government was prepared to give Citi when it was trying to buy Wachovia. In that case – which was the model for this bailout – the state was to receive $12bn of preferred shares and warrants.
What’s more, with the Wachovia deal, Citi would have had to absorb the first $42bn hit. This time the first loss is only $29bn although Citi would also get hit with 10% of any subsequent losses. Given that the size of Wachovia’s portfolio was $312bn – almost exactly the same size as Citi’s own bad bank – Pandit looks to have got a better and cheaper insurance policy.
Of course, Citi may not have got a better deal overall in that, with Wachovia, it was also going to get a large deposit base at a knock-down price. What’s more, the story isn’t entirely over because the $306bn valuation of Citi’s bad bank hasn’t yet been agreed by the government. If the Treasury concludes that a lower value should be used, that would give the taxpayer a bigger buffer against losses while shifting more of the pain to Citi. Taxpayers should hope that Hank Paulson, the Treasury secretary, negotiates sensibly on their behalf. As for Pandit, he should wait before cracking open the champagne. He knows too well how easily deals can slip through his grasp.