The WSJ’s point was that the irregularities could have been the result of banks trying to mask their risk levels in the past five quarters by temporarily lowering their debt, just before reporting it to the public.
It’s interesting then that Barclays Capital’s Joseph Abate should point out in a recent note that there’s still not been much respite in the quarterly anomalies:
More interestingly, there is a very pronounced and recurring seasonal pattern in the volume of repo trades at quarter-end. Since 2008, in the week before the quarter-end, volumes typically plunge by 5-15%. Most of the decline occurs in the term market where the quarter-end drops are even more pronounced (around 25%). Following the plummet at quarter-end, it normally takes an additional eight weeks for activity to recover to prequarter- end levels.
Although the quarter-end trend is not new, it has recently drawn the attention of regulators – particularly as the primary cause is bank reporting pressures. Since banks report a quarter-end snap shot, they have a strong incentive to lower their (term) repo balances to appear less leveraged. While it is not clear what might result, a number of efforts are underway to rein in repo market activity, including the FDIC’s efforts to impose a haircut on secured creditors, as well as the Fed’s triparty repo reform initiative.
In fact if you look at the data, it seems the discrepancies may have gotten even more pronounced (if anything).
First, here’s Abate’s analysis showing an increase in the ‘total repo outstanding’ as the end of the quarter draws near:
And here are a couple of graphs straight from the raw NY Fed data:
Worth thinking about.
Repo is still puzzlingly inverted – FT Alphaville
The Repo 18: It’s Not the Collateral, It’s the Cover-Up – Baseline Scenario