News 
Market Data 
Discover
Replies to post #9 on NSS ~ Counterfeiting Stock ~ MM Games Played
07/24/10 9:20 PM
Mr. Byrne:
I notice that Overstock.com has not been on the Reg SHO list for some time, and that few stocks are on it now. Does that mean the rules have worked, or what other comments would you make?
Floyd Norris
Dear Mr. Norris,
Thank you for asking. As you may know, I have never primarily considered this fight to be about Overstock: it has been a fight to curtail illegal-but-winked at stock manipulation in our marketplace before more companies were destroyed and our marketplace destabilized. I always felt the insistence by some that my fight was about Overstock was part of a cover-up, and so I commend you for asking regarding the greater picture.
If I learned that the percentage of HIV-infection in the public blood supply had dropped, it would give me cheer, but I would not say, “This level of HIV infected blood has reached an acceptable level.” Similarly, seeing this drop in the number of firms on the Reg SHO list and the number of failed shares brings me cheer, but does not make me think the game is won. In particular, until we have data on failures in the other nooks and crannies of the system (broker-level netting, pre-netting, Stock Borrow Program, ex-clearing and off-shore failures) I cannot celebrate victory. However, yes I do think that the rule changes put in place during the latter part of last year made stock manipulation harder, and people have to be more sophisticated to do it.
Ultimately, what has to happen is that we go to requirements for pre-borrow and hard-delivery. That is what it will take to really end things. Incidentally, that is where the discourse was some months ago when it log-jammed within the SEC: on one side, the good guys arguing for pre-borrow and hard-delivery; on the other side, SIFMA and the Managed Funds Association arguing against them (on the grounds that if they were forced to stop doing that thing they have all along sworn they were not doing, it would dramatically affect liquidity). I believe (and have a source from within the SEC who tells me) that the current debate over the uptick rule was largely manufactured to shift the discourse. It is a red herring.
Respectfully,
Patrick Byrne
Thank you for your prompt reply.
Is there evidence available on the scale of the “failures in the other nooks and crannies of the system (broker-level netting, pre-netting, Stock Borrow Program, ex-clearing and off-shore failures)” that you refer to? Netting sounds as if it removes the possibility of manipulation. Is there a reason to think otherwise?
Floyd Norris
-----
Dear Floyd,
First, just to make sure you know what I am talking about, I am sending you this section from an essay I wrote last August 3rd:
Let us look at where these unsettled trades can reside within the piping of the “factory” that is our nation’s stock settlement system, The Depository Trust & Clearing Company (“DTCC”). I will use Goldman and Morgan as hypothetical examples only.
“Desked trades” – Imagine Goldman takes your order for 1,000 shares of stock, but stashes your order in a desk and sends you statements saying that you have those 1,000 shares in your account (and use your money towards the $10 billion they pay themselves at the end of the year for being so clever). They have written a CDF to you without your knowledge: there is a 1,000 share failure-to-deliver to you at Goldman (which no one else knows about, incidentally).
“Pre-netting” – Goldman has one client sell 5,000 shares and another buys 3,000. The seller never delivers. Goldman “pre-nets” the trades before submitting them to the DTCC. Hence, the DTCC sees only 2,000 shares of the failure.
“CNS netting” – Goldman submits to the DTCC’s Continuous Net Settlement system that it sold 2,000 shares that it does not deliver. Imagine Morgan Stanley was on the other side of that particular trade. But maybe Morgan has a client who sold 1,000 to a Goldman client, and which that Morgan client failed-to-deliver. The DTCC nets the two trades, and therefore sees just 1,000 shares of failure (Goldman to Morgan).
“Stock Borrow Program” (“SBP”) – The DTCC looks at that 1,000 share failure, and says, “We have 400 shares we can loan Goldman from our Stock Borrow Program”, i.e., from the accounts of other BD’s within the DTCC. That reduces the failures it sees to 600.
“Ex-clearing” – Suppose Goldman and Morgan apply to the DTCC to move 500 of those fails ex-clearing, and the DTCC approves. Those 500 FTD’s are turned into a derivative contract between Goldman and Morgan. As a private contract, it is not regulated by the SEC, and the DTCC does not even know when that contract gets cleaned up, if ever.
“Offshore Failures” – Suppose someone sells 1,000 shares into this market from a foreign offshore exchange? There is a different terminology to describe such failures, and therefore the data is hard to get to. What is clear, however, is that there is little pressure to clean up failures among exchanges.
In this example, there are 100 failures at the CNS level. Yet there were 7,000 failures throughout the system. Therefore, we should remember that, however many unsettled equity trades there are at the CNS level, it is likely to be a fraction, and maybe a quite tiny fraction, of the total unsettled trades in the system.
What is the ratio of total fails in the system to those trapped in the CNS system? No one seems to know (and in fact, while the individual pieces of data are known individually, I strongly suspect that no one party has the bird’s eye view of how many of these there are at all levels). The estimates I am told range from 3 to 15. For ease I will refer to this as, “The Iceberg Principle” and the ratio of total failures to CNS failures as “I”.
So how big a problem is this?
·The last reported size of the failures-to-deliver at the CNS level are $8.7 billion.
·By Iceberg Principle, total failures = I X $8.7 billion ˜ $30 to $120 billion.
·By Feynman Principle, total cost to cover = F X I X $8.7b = F X ($30 to $120 billion).
So respectfully, Wall Street, I believe you are Oak Ridge, Tennessee, blithely going about your jobs at the factory, taking for granted “the piping” that is our settlement system. I believe you have manufactured, and are sitting squarely on top of, a financial atomic bomb. That’s not good for you, of course, and if it goes critical, America is downwind.
Second: So now your questions are: Doesn’t netting remove the possibility of manipulation? And, What evidence is there regarding the size of failures in those crevices?
a) Netting does not remove the possibility of manipulation, but masks it instead. Assume Goldman has a client selling 1,000 shares of stock ABC, and Morgan has a client who buys those 1,000 shares, but the shares never deliver. However, Morgan has another client who sells 1,000 shares, and Goldman has one who buys those 1,000, but they also never deliver. In total, there have been 2,000 shares sold but not delivered. However, by pre-netting the brokers’ fails, those failures disappear. Collectively the two sellers have weighed down the market with 2,000 shares of phantom stock, but the netting makes 0 fails show at the DTCC.
b) What is the evidence supporting my claims regarding the size of the failures in those other crevices? I am 100% confident of my claims because I possess incontrovertible proof that, for one particular stock whose name for legal reasons I cannot disclose (but you may be able to guess), the failures maintained by justone broker injust the ex-clearing crevice have been, at times, greater than the sum total of the failures showing up in the CNS system: Thus to me, this is a settled question.
For one not in a position to review such evidence, however, you will have to do what I have done, which is, interview people in the settlement industry, find out where trade failures can persist, and for how long, and why. I think any fair-minded researcher will quickly be told of how much slop there is in these stock settlement niches, how easy it is to take things ex-clearing, how deep the offshore failures run, and so forth. Then the fair-minded researcher will try to get some data from the SEC, or DTCC, and will discover that he is stonewalled on even the most basic information, a fact which would, I think, raise the suspicions of that fair-minded researcher. Indeed, this is where some members of your esteemed profession have rubbed me raw, because instead of saying, “Why can’t the Establishment release the data, if there is really nothing there?” (as until a few years ago I would have expected any journalist to reason), instead almost all of them have said, “Well, I agree there is evidence of settlement failures, but the Establishment says there’s no problem, and they won’t release the data to show me one way or another, so I’ll just stop digging.” After all, remember that we are not debating the properties of some newly-discovered subatomic particles: this data is knowable, and is in fact, known. Good luck trying to get anyone to release it to you (without filing the kind of massive, one-in-a-lifetime lawsuit against an entire industry, as I have done).
I hope this helps.
Patrick
