The biggest surprise in JPM's Q4 earnings release was not the firm's legal troubles: those are well-known, and largely priced in even if JPM did generously add back 27 cents in EPS to the adjusted bottom line, which means that if JPM were to treat its legal expenses as recurring (as they have been for two years now), its non-GAAP EPS would have been $1.13. No, the biggest surprise by far was that as of this quarter in addition to its trusty use of DVA or a Debt Valuation Adjustment (the old fudge when a bank "benefits" when its credit spreads blow out) JPM also added the use of a Funding Valuation Adjustment or FVA.
The amount of the FVA benefit? A whopping $1.5 billion addback to GAAP EPS, which together with DVA, resulted in a $2.0 billion pretax loss, promptly added back to get boosted non-GAAP EPS (and recall $1.3 billion in GAAP JPM "earnings" came from reserve releases).
Here is how JPM's explained the adoption of FVA as a bottom line fudge:
[8]
The punchline: "For the first time this quarter, we were able to clearly observe the existence of funding costs in market clearing levels"
For those not familiar with FVA, here is a refresher from Risk [9]:
Banks that include a funding valuation adjustment (FVA) in derivatives prices may be vulnerable to predatory customers, according to two academics – a claim that is already being attacked by traders.
* * *
FVA reflects the costs a bank incurs when hedging an uncollateralised trade with an offsetting position on which collateral is required. When the former is in-the-money for the dealer, it will not receive any collateral from its customer and would have to fund its own posting to the hedge counterparty. Many dealers recognise FVA by discounting uncollateralised trades at their own cost of funds, meaning banks with a higher cost will charge a lower price for trades that create a funding benefit – a disparity Hull and White claim customers can exploit.
"It's a perverse situation. The FVA is determined by discounting derivative payoffs at your own funding rate, which lowers the price. If your funding cost is sufficiently high, your valuation will be below the market clearing price. Banks can then book mark-to-market profits from the difference – a form of accounting money machine. Some dealers with high funding costs appear to be doing this now, but if they have to reverse the accounting treatment it will result in very large writedowns – heads will roll," says White.
The paper claims the arbitrage can be achieved by buying an option from a bank with a higher funding cost – translating into a lower price because the dealer will recognise the benefit generated by the upfront option premium – and selling the same option to a bank with a lower funding cost. Because of the spread between the two banks' prices, each can be offered a small premium above the FVA-inclusive value by the client, which pockets the difference remaining after also hedging its counterparty exposure to the bank with the higher funding cost. Each bank appears to have made a profit relative to its FVA-adjusted price. However, according to Hull and White, because each bank's hedged portfolio will only earn what they refer to as a risk-free rate, the bank with the higher funding cost has priced the trade too cheaply and will actually lose money.
"This should be giving people pause for thought," says Hull. "I have spoken to one end-user – I won't name names – and it is seriously considering this. It may take a while but eventually dealers will realise they are on the wrong end of this and correct their prices. If Microsoft – to pick a name at random – can get different prices from Royal Bank of Scotland and JP Morgan, they will look to exploit it."
Since the arbitrage would only be available on uncollateralised trades, it appears to be restricted to the shrinking group of clients that enjoys this privilege – typically sovereigns, supranationals and agencies, corporates, special-purpose vehicles and some large pension funds – none of which are traditional arbitrageurs. However, Hull claims others, such as hedge funds, which are usually required by dealers to collateralise their trades, may be able to access it through ingenious financial engineering.
"Hedge funds employ a lot of smart guys who are paid a lot to spend all day thinking up ways of making money. I expect there will be people out there now looking at how to exploit this. I've been in the derivatives business for 35 years and I've seen this kind of thing happen time after time. Someone will do it," he says. One way might be for a hedge fund to implement the trade via a corporate, Hull suggests.
Bankers are not convinced, citing what they see as flaws in the argument, such as the existence of a risk-free rate – considered outmoded in post-crisis markets. They also argue there are few customers, if any, that would be able to take advantage of the arbitrage, because it requires a proprietary trader that has no collateral agreement with its dealers and routinely hedges its counterparty risk.
"It's not an arbitrage, end of story. You have credit risk to the higher-funding-cost bank and you can't be so blasé about wishing it away. That costs money and will eat into any profits. And I got a bit angry when I saw them talking about the risk-free rate. There is no risk-free rate. You discount a collateralised trade at the overnight indexed swap rate because it's the rate that funds it, not because it's risk free – it's not," says one emerging markets trader at a European bank. "They keep saying things like ‘prices should not include FVA'. Well, OK – but they do. If they didn't, the banks would go out of business. It's nice theory, but it does not reflect reality."
A senior treasurer at a European bank echoes these objections: "This is turning into a religious debate now"...
Religious or not, as of this quarter even the almighty JPM is subject to collateral funding costs. For now, it is an addback. Let's see what happens when it has to subtract from the bottom line...