UPDATE 8-13. NIPC has drafted a comment letter to the SEC on its umpteenth deliberation on eliminating the market maker exemption to REG SHO. You can download it here. It spells out exactly why the SEC's recent emergency action is a load of hooey, because if the SEC enforced the existing laws on the books, there would be no need for new emergency actions. It also spells out why REG SHO's supposed crackdown on FTDs is a load of bullocks. What the SEC really should do is remove the locate loophole and require a hard borrow, and eliminate the abusive market maker exemption - then it wouldn't need to do anything more but enforce the laws already in place. Instead, it has created contradictory and more liberal rules, a la SHO. Why the hell is it so difficult for the SEC to just enforce the federal securities laws, instead of doing decade-long contortions to avoid doing so? Fair question. Afraid I already know the answer. And it has resulted in a financial system that is moments away from running headfirst off the cliff.
I would suggest any and all to add your own comment letters to the SEC, demanding that the SEC respond publicly to the very reasonable sets of questions in the letter. Wanting to understand why the cops aren't enforcing the law is a fair inquiry, I think...
--------------------------------------------------------------------
How big is the problem? I mean, we hear the now famous $6 billion delivery failure number tossed around from the DTCC, but how accurate is that, really? Is it a complete answer? Is there more information that is knowable?
The answer is, yes, more is knowable.
"Approximately US$1.8 trillion worth of trades remain outstanding and unsettled globally every business day, contributing significant credit and operational risk exposure to the trading participants."
That from the document at Touchbriefings.com. Huh. $1.8 trillion is a big number, even by Pentagon standards.
http://www.touchbriefings.com/pdf/1417/kumar.pdf
Or how about this? From the DTCC:
"For example DTCC estimates that 5% of secondary market trades fail to settle each day. With approximately $4.5 trillion of settlement value in 2004, failed transactions equal $ 225 billion daily."
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=849224
Alternatively, we can go to the UK Exchange Handbook, which contains some fascinating perspective on our wondrous settlement system.
http://www.exchange-handbook.co.uk/index.cfm?section=articles&action=detail&id=38756
If I read it correctly, it says that at least 5% of DTCC transactions fail to deliver. It also hints at delivery failures because of the CDS direct cross border link into the DTCC and between the prime brokerages.
"One of the problems in assessing how reliable, and hence safe, a market settlement system is concerns obtaining reliable data on such matters as failing transactions -- let alone ascertaining exactly where liabilities lie during the settlement and subsequent on-going custody of the assets. In the more advanced markets, such as those in the UK and the US, the local regulators have ensured that reliable transaction data is readily available in a very transparent manner. In these markets, when the depository advises that they have a fail rate of approximately 5% (in the case of DTCC) or approaching 1% in the case of CRESTCo in the UK, one can rely on such figures."
"This research is supported by the DTCC white paper, which reports that 6% of institutional transactions being settled on an average day are expected to fail, while the fail rate on the market-side is lower at 4.4%."
Huh. Very interesting. Again, sounds larger than that $6 billion per day, which many assume is a rolling total. I mean, 5% of everything they process fails? After netting and the stock borrow program have minimized the number that could fail? That is an amazingly large number, 5 times larger than what happens in Britain. Scary part there is that we are now talking after the replenishable pool from the stock borrow program is used, and presumably massive numbers are shunted ex-clearing, and not counting desked trades, and after CNS netting handles the vast majority, we still have 5%? Yikes. Makes me start to believe that the true number is many multiples of that 5% number. Anyone with hard data that can show that assumption isn't true is welcome to bring it on. Facts are facts.
So then, in our quest to understand it all, we turn to the actual white paper published by the DTCC on settlement issues, and we get more data. Much more.
http://www.dtcc.com/downloads/leadership/whitepapers/settlement.pdf
"FAILS AND RECLAIMS
This shortfall in effecting STP is largely an efficiency issue. Fails and reclaims are another matter. They create risks for participants and for the system as a whole.
Because fails are quite common in today’s system, it falls far short of straight-through processing. Currently about 5% of trades fail or are “dropped” at the end of the day — about 20,000 from CNS and 15,000 from non-CNS deliveries for a total of about 35,000 of the typical day’s 700,000. This doesn’t include “fails to deliver” that aren’t even introduced into the system, which would make the fail rate higher (that's ex-clearing they are discussing, BTW - Bobo). Fails create significant risks for the deliverer and receiver. When a fail occurs, the deliverer is short of funds (although a firm can lend the securities it has and replenish them in normal market circumstances). Both the deliverer and the receiver have portfolios that are not what they expected. Although a fail does not create a credit, counterparty or principal risk, it does create a liquidity risk for the deliverer. And institutional trades that fail create position risk for both deliverer and receiver.12
Most fails occur because positions are not available; that is, the deliverer does not have free inventory. Stock lending can, of course, overcome this, as long as the lent stock is available quickly enough. When, as is usually the case, the lender receives sufficient cash collateral from the borrower, the credit risks associated with stock lending are small compared with the benefit of eliminating settlement fails."
Huh again. So, 5% fail PER DAY, not counting ex-clearing. Not a rolling number, but per day, unless I am reading this wrong. I don't think I am, but it is always possible, and I would welcome brighter minds than mine shedding light on this if I'm getting any part of it wrong.
So, this is a big, big problem, and it isn't happening because the dog ate the certificate. It's happening because the brokers just flat out don't have what they sold. As in, fraud. No got the doggee in the window I took the money for.
How big is the total problem including ex-clearing? Nobody's telling how large ex-clearing is, but we have some hints that it is much larger than the in-system fails. If so, we are talking systemic meltdown sort of size, and it underscores why the industry fights so hard to pretend that it is all a minor issue, or lunacy from fringe kooks.
Kinda tough to argue that with the above citations, isn't it? Oh. Those.
So the turd brigade posts like mad over at the Standard, hoping to remove any semblance of intelligent thought from the discourse. They accuse Patrick of being a cheat or a loon, they accuse me of being a stock manipulator or crazy, but what they don't do is cite hard fact to support their ramblings. Nothing new there. They think we are all stupid. They are the masters of the universe. We are the sheep. Problem is, every now and then the sheep figure it out, and it makes stealing everything the sheep own harder, which they don't like. Because all thieves and criminals prefer easy money - that's why they are thieves and criminals, versus productive members of society.
But I digress...
Special thanks to DavidN for digging these up.