Would you agree that the "risk" associated with the "pricing" of a matched order is qualitatively different than the "risk" associated with not making the fill at all...
Yes, they are different types of risk, with a common denominator: they both, in their worst version, result in losing institutional calls. Lose enough calls, and you're out of business.
...in a way that might make the "risk"...that the eVWAP price is "inferior" somewhat more tolerable for an institutional trading party that might otherwise eschew electronic trading approaches altogether...
It depends on the institution, and that's where any attempt to connotate or otherwise suggest universal applicability for this - or any trading system - fails.
And offhand, no; I don't see any way in which getting a block of $20 bid stock for $19.75 (whether in whole or part) at the end of the day is better than getting it for $17 at noon.
...especially where the task at hand for the institutional customer is filling a huge order that will require a variety of methodologies to fill, and where there is some recognition that parts of such a huge fill made the "regular way" will still necessarily be at "inferior" prices compared to other parts of the same block?
Not at all. The larger a block is, the more insistent the sellside desk is that it gets done in one piece. The reason being, of course, because it's one price, one print, and can be controlled easier than 20 pieces being done all over the place. One shot, one kill.
An excellent, while superlative, example of this is the British Petroleum block traded by Goldman Sachs in 1997 - at that time, the largest block ever traded. The Kuwaiti Investment Office sought to sell 170,000,000 (one hundred and seventy million) shares worth roughly $5 billion dollars, which was bid upon by several firms and Goldman won. The price at which the stock was trading in the market that day in London was 744 pence/sh ($1,066/share); their winning bid was 711 pence/sh ($1019/share), as I remember hearing. One price for the whole thing, and it's done.
The next concern, of course, is placement.
Do you think the that the risk associated with getting a "true VWAP price" which may or may not be optimal...is qualitatively different...than...waiting for "information flow relating to a stock" and seeing if "something in the market" presents a fill price "far beyond the scope of VWAP, RPM, or any other measures" might actually lead to inferior pricing.
In terms of the risk identified by opportunity cost (and again, the risk to my future of calls with this buyside firm or fund) in a liquid stock, my intuition would be that it's far less risky to see what's going on in the market than to use a system such as the one we're discussing.
If the market goes against me, I can always take a hit and make the price, so long as I have or can get the stock. That is the flexibility we spoke of earlier - the ability to make prices, as opposed to taking prices.
Mistakes can, and do, lead to sellside traders printing transactions for their buyside customers at a loss to themselves. That, again, is part of the trading relationship concept.
The second generation of eVWAP is going to be what they describe as a "semi-continuous" market, where parties can enter the system at 15 minute intervals and match with each other, with fills to be made at the volume weighted average price calculated during each 15 minute interval.
Sounds like they're stepping into OptiMark territory here, which in my opinion is not a good move. To match successfully once a day, when there is some uniqueness to a system, allows some liquidity - the critical mass toward more liquidity - to build. To now break that down into sessions during continuous market hours...I suspect might result in an overall drop in volume. At the least - again, as per my intuition - maybe the same numbers for the end-of-day cross, but little volume above a certain plateau intraday.
if eVWAP gets utilized...by 3 or more of the world's largest money managers and 100-200 other large institutions and B/Ds, that problem may well be overcome...
Huge institutional participation, or at least the promise of that participation, didn't help OptiMark, and it hasn't helped other highly anticipated systems, either. And when you cite these large buyside entities, you must remember that they are major participants of such practices as directed brokerage, commission recapture, and the like, few of which are adequately addressed by electronic systems at this point.
...and the advantages of having an anonymous, secure... and ...where the presence of the particular large block being traded is unknown to the counterparties...
OptiMark had a Big Five firm certify the integrity of its' anonymous, secure matching system, and it folded all the same. Not a selling point; its the bare minimum of service that you expect, buy or sellside.
dynamic trading venue
Well, like I said - intraday sessions are not the way I'd go, for several reasons, but as this has yet to develop, it may or may not prove "dynamic." I'm skeptical. In my humble opinion, there are other, far better ways they could increase the utility of this product and differentiate themselves than by taking on POSIT, the AZX, LiquidNet, and every block trading desk out there during market hours.
Thinking out loud, now: I can already envision one possible result of eVWAP sessions intraday.
Knowing, as we all will, when they are and therefore watching for when the prints hit the tape, block positioners and sales traders will simply say, "see when the eVWAP print comes, and whatever it is, we'll give you a $x per share less."
Done! E-VWAP essentially prices, and retroactively permits itself to be effectively underbid, every single time - in the minutes following a crossing session. Realized, this would actively route trading business to competing, human staffed trading desks.
What do you think of the effect of decimilization on your job as a trader trying to fill large blocks?
On trying to price blocks, it's great. More price points, easier to clear, etc. On trying to trade blocks, it's tricky. One must remember that lots of these folks have legitimate complaints, and some are just afraid of change.
I have long said, on SI and amongst my colleagues, that going to nickel increments would have been smarter first, following up with either two cent or one cent intervals, but, you play the cards you're dealt. :)
LPS5 |
