VWAP vs BSP

[showtrader]

I'm trying to back all runners which have a BSP which is X% greater than the VWAP.

Of course the issue is that it's hard to know the exact BSP prior to the off, however I always thought LTP or a midprice (stored value) would be a pretty good approximation.

However, I'm finding that even if I trigger a bet at 1000 (and then set to take SP) if LTP is > 1.X x VWAP, I'm not actually backing the same horses that meet the same criteria when I check the BSP and PPWAP figures that Betfair publish the next day. Most of the ones I'm backing do meet the criteria, but there's actually around three times as many which fit the criteria which I don't end up backing.

Does anyone know why this might be? Does the Bet Angel VWAP differ from the Betfair published PPWAP in some way? BSPs don't generally differ than much from the LTP, so I'm unsure what would lead to such a big difference in which bets meet the criteria.

Any ideas?

Thanks

[Silverthorn]

VWAP = Volume Weighted Average Price NOW so keeps updating during the race
PPWAP = Weighted Average Price of all bets placed pre-off so does not update during the race

The same until the jump but totally different once the trade goes in play.

[showtrader]

Yeah - so they should be the same pre-race up to the off then, I'm not betting in play, I'm taking BSP.

So any ideas why I end up only backing about a third of the selections which match the criteria when I check the Betfair data the next day? Are there really that many runners which jump that much in price at BSP vs what their LTP was just before the off?

[brodandre]

Apologies for being a bit late to the game on this thread, but I’ve done some looking into this independently and have formed some theories on my own betting that might be at play here…

Two questions: have you compared the last recorded VWAP numbers to the PPWAP? If so, are the PPWAP numbers consistently skewing closer to the actual BSP than the last recorded VWAP you saw in BA? In some of the markets I’m bettinf, the actual BSP ends up being shorter than my estimates pre-off, and the VWAP moves markedly in that direction as well, which I reckon means there’s likely a substantial weight of money on the back side going at BSP in comparison to the pre-off non-BSP traded volume. I still have more digging to do on my specific cases but perhaps this or a similar circumstance is at play with your selections?

[decomez6]

I'm trying to back all runners which have a BSP which is X% greater than the VWAP

Do you round off your vwap calculations in BA? How many decimal places ?
checked gutuami’s approach on calculating to one decimal place
( vwap multiply by 10 then round to the nearest whole number)

Of course the issue is that it's hard to know the exact BSP prior to the off, however I always thought LTP or a midprice (stored value) would be a pretty good approximation.

On how betfair calculate BSP


https://docs.developer.betfair.com/plu ... 92566

[Kai]

VWAP = Volume Weighted Average Price NOW so keeps updating during the race
PPWAP = Weighted Average Price of all bets placed pre-off so does not update during the race

WWPD = What Would Peter Do so a very tricky one to calculate
MMMBop = A Hanson song released in 1997 so does depend on musical taste

Apologies for the thread intrusion, no derailment intended, carry on

[sniffer66]

VWAP = Volume Weighted Average Price NOW so keeps updating during the race
PPWAP = Weighted Average Price of all bets placed pre-off so does not update during the race

WWPD = What Would Peter Do so a very tricky one to calculate
MMMBop = A Hanson song released in 1997 so does depend on musical taste

Apologies for the thread intrusion, no derailment intended, carry on

[deestar]

Yeah - so they should be the same pre-race up to the off then, I'm not betting in play, I'm taking BSP.

So any ideas why I end up only backing about a third of the selections which match the criteria when I check the Betfair data the next day? Are there really that many runners which jump that much in price at BSP vs what their LTP was just before the off?

In short the answer is yes. They can be quite different. I have a similar churn strategy for PC purposes but on the lay side. If your idea is to simply back all BSP that are greater than x % of the PPWAP pre off because the historical data shows doing that is +EV; I'm sorry but it ain't that easy.

PPWAP includes (I believe) amounts matched at BSP so that will skew the data. Your own BSP bets will impact the BSP particularly with long odds runners with low volume.

If you think you are on to something why don't you take BSP out of the equation and simply back when the current price is x% greater than the VWAP at a certain point pre-off?

[decomez6]

VWAP = Volume Weighted Average Price NOW so keeps updating during the race
PPWAP = Weighted Average Price of all bets placed pre-off so does not update during the race

WWPD = What Would Peter Do so a very tricky one to calculate
MMMBop = A Hanson song released in 1997 so does depend on musical taste

Apologies for the thread intrusion, no derailment intended, carry on