Derivatives datapoint of the day, OCC statistics edition
" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">
Consider this, from the executive summary:
Five large commercial banks represent 97% of the total banking industry notional amounts and 88% of industry net current credit exposure.
What are those five large banks? According to the OCC, they’re JP Morgan Chase, Goldman Sachs, Bank of America, Citibank, and Wells Fargo. Add up their “total derivatives” numbers in Table 1 of the latest OCC report, and you get $197 trillion, which is indeed 97% of the $203 trillion in total notionals.
But never mind that, and take a look at Table 2 instead. That shows the notional amounts at bank holding companies, rather than banks themselves. Suddenly, the size of Bank of America’s derivatives holdings spikes from $39 trillion to $75 trillion, while Morgan Stanley appears from nowhere to reveal itself as holding a more-than-healthy $41 trillion in derivatives. It seems that at Merrill Lynch and Morgan Stanley, the derivatives are generally held by the holdco rather than the bank, which allows the OCC to ignore them for the purposes of its headline calculations.
Add up the derivatives books at the holdcos, and the total isn’t $203 trillion any more: it’s $291 trillion — an increase of $88 trillion which is very hard to find in the OCC report unless you’re specifically looking for it. And never mind Wells Fargo, which was also something of an also-ran in the top five banks. The top five now comprise JP Morgan, Bank of America, Goldman Sachs, Morgan Stanley, and Citigroup, with derivatives holdings between them of $278 trillion. That’s 95% of the true total, or 137% of what the OCC would have you believe was the total.
Why does the OCC make the rather legalistic distinction between commercial banks and holding companies, and then concentrate on the former rather than the latter? If the object of the exercise is to add up total derivatives exposure, that doesn’t make any sense. Someone there should really start asking themselves why they’re publishing this report, and how they could make it most helpful to the people reading it.