LPS5 - I think maybe you are misinterpreting Rob's question, which may be more strictly mathematical in its orientation than the overtone within it that you seem to be responding to.
If I understand Rob correctly, his point is predicated on the notion that traders will use typically market orders to fill (whether as a strategy because the fill is more important at the moment than the appearance - perhaps illusion - of price improvement derived from sticking with a limit order, or otherwise). If you assume that buyer's market orders fill at the ask, and seller's market orders fill at the bid, and that the VWAP price effectively averages between all executions (which are carried out on that basis) throughout the day, doesn't the pure derivative (VWAP) ultimately have to represent price improvement against those who time slice perfectly, since the time sliced trades (at least if made with market orders) will always be on the wrong side of that spread with each trade? So, in effect, doesn't the regular way trader have to outperform a true average of market orders staggered throughout the day to approximate volume activity to stay even with the pure VWAP price given the effect that the spread alone has on the derivation of the average price? I suspect my answer (if I were answering this one for you) might be that the strategy of deciding when to fill ahead of a movement of momentum brought on by the impact of an influx of buying, i.e., increasing the volume of your order rather than adhere to a pure time slicing approach -- because you perceive your presence and that of other buyers will drive the price up making an earlier fill advantageous (so that upping the order ahead of schedule will result in real price improvement) -- is what separates the great traders from the average ones. But with the Elkins McSherry study showing that 83% of the time, the fills do not "beat the VWAP", I suspect that "great traders" are not exactly a dime a dozen and that use of a passive approach which, when it matches, achieves an average that statistically represents price improvement is not a bad thing, and may fit it as a component of an active trading strategy.
It seemed your responses to Rob were more focused on whether, as a tool, a system which matches trades made blindly and passively at a passive benchmark derivative price is necessarily achieving what a trader is asked to achieve by the portfolio manager. And I totally agree that the traders priority should be, first and foremost, to meet the requirements established by the customer, whatever they might be. Which may be, among other things, to play a more active role in the fill than a passive system would appear capable of achieving.
You mentioned in one of your earlier posts here something to the effect of how placing ones order into a server and waiting to see how it turns out is somewhat antithetical to the notion of a trader actively working the customer's order to ensure its requirements are met. The difference with eVWAP, it seems to me, is that you must place your order before the open, and are advised as to the size of your fill on eVWAP, at least 10 minutes before the market opens. The only thing you "wait on" is the price, not the fill. You have the entire trading day to work the rest of your order. So why wouldn't a trader, knowing that it has a 400,000 share order to fill, see if it could not succeed in filling some percentage of it at the VWAP price and reduce the market impact of the rest of its order? It seems like a no lose proposition, particularly if the Elkins, McSherry data is correct.
MST |
