Price Leverage - to Lay to a lower price than is available

[Fixador]

It sounds like 'Hedging' - but it isnt - is it ?

With a view to getting a lower Lay price, i stumbled onto “Price Leverage”. But my poor ol head is struggling to get around this - here is an example, taken from BF education

Principle : Buy HIGH with small sum, then SELL low with larger sum.

Example :-

(1) Back runner @3.5 with £20 stake

(2) Lay runner @2.3 with £100 stake

Net situation equilvent to LAY @ 2.0 with £80 stake

Do you understand the math ? cause i dont ! And as for the bailout if the price drifts - no idea !

Please help !

[PeteB]

ok it breaks down like this...

backpnl = round(backstake * (odds - 1), 2)
laypnl = round(laystake * (odds - 1), 2)

ifwin = backpnl - laypnl
iflose = laystake - backstake

equivalent odds = 1 - ifwin / iflose
(though this only really works if one of ifwin or iflose is -ve - if they are both +ve or both -ve it's a bit meaningless)

So let's say we have placed the back £20 @ 3.5 and it's been matched

backpnl = 20 * 2.5 = £50
laypnl = 0
ifwin = £50
iflose = -£20
equivalent odds = 1 - 50/-20 = 3.5 (ok!)

Now the price has fallen and we can now lay at 2.3 (lucky us) - how much should we lay??

To get a free back, we want iflose to be zero - i.e. backstake = laystake. So we stake the same amount - we lay £20 @ 2.3
backpnl = 20 * 2.5 = £50
laypnl = 20 * 1.3 = £26
ifwin = £24
iflose = £0
So we've got infinite value, but we will only get any money if it comes in

To get a free lay, we want ifwin to be zero i.e. backpnl = laypnl. So our stake is backpnl / (odds -1). So we lay £50 / 1.3 = £38.46 @ 2.3
backpnl = 20 * 2.5 = £50
laypnl = 38.46 * 1.3 = £50
ifwin = £0
iflose = £18.46
Again we've got infinite value, but only get any money if it loses

To be green, we want ifwin = iflose. So we stake backstake * backodds / layodds. We lay £30.43 @ 2.3
backpnl = 20 * 2.5 = £50
laypnl = 30.43 * 1.3 = £39.46
ifwin = £10.44
iflose = £10.43
Now we have guaranteed free money

And for their example
backpnl = 20 * 2.5 = £50
laypnl = 100 * 1.3 = £130
ifwin = -£80
iflose = £80
equivalent odds = 1 - -80/80 = 2
I.e. we have effectively laid £80 at evens

So which of these do we do??
Well, if we have an estimate of the real probability of the selection winning, then we could work out the expected return
expectedreturn = ifwin * probwin + iflose * (1- probwin)
and then pick the one with the highest expected return

But we also need to know what our staking plan is, and how long a losing sequence we can handle. If we green up or take the free back or free lay, then losing sequences don't come into it - we have no risk of ruin. If we take the price leverage option, then we risk losing £80, and so it will only take a losing sequence of £bank / £80 bets to blow the account, and not be able to play any more. As in any game with a small edge in your favour (assuming you have this) the most important thing is to retain the ability to keep playing!

So when you have the option of reducing risk of ruin to zero, it seems crazy (to me) to take the price leverage option and still face a risk of ruin.

This aside, you can then chose whether to take green, free back, or free lay (or somewhere in between) based on the expectedreturn calculation...

[PeteB]

It's also worth saying that in the free back, free lay, or green scenario, I get my £20 back as soon as I place the lay, and I can now play with it again elsewhere. In the price leverage scenario, I have had £20 tied up for while, and I now tie up a further £60, further reducing my ability to play elsewhere until the market settles.

And I forgot to factor in commission - if you include this, then you need to have very strong opinions about the result (i.e. you know it is fixed) to make it worth having any significant difference between ifwin and iflose.

Still people will use the price leverage, in the same way that people use loss-chasing staking plans without calculating their risk of ruin...

[Fixador]

PeteB : That has got to be ( for me - at least ) one of the most illuminating posts i have ever read . I am going to be studying that for a week if not 2 or 3 ! ie, it has arrived just at the right moment on my learning curve.

My query arose - as a kind of detached personal challenge - in response to what i considered to be a dubious set of numbers - yet could not prove immediately. Basically a set of selections to Lay , but with historical data given at SP ( alarm bells started to ring ! ).

I took the methodology to Massey's database - to compare published performance , with Massey's database for the same period. The %WinReturn to SP correlated - 80% with a 20% strikerate ( if BACKing )

I then used the methodology , for earlier periods of data in Massey's database - and oh dear ! the methodology plummented in performance - the %WinReturn = 90% and strikerate still 20% ( if BACKing ) consistently from 2004 to 2007.

So , now the methodology begins to resemble a statistical outlier ! ( More alarm bells ! )

One for the bin !

However, I was left curious by the some of the numbers:-

What does a BACKing %WinReturn of 80% to SP with a strikerate of 20% ( if BACKing ) mean, if attempted to Lay a methodology with this performance:-


Very roughly : %LayReturn to SP = 100 % - 80% = 20% gross

But who can Lay at SP ? so deduct 13% for an average Lay Odds ( SP +13% ), and 5% for commission

%LayReturn = 100 % - 80% -13% - 5% = 2% -------------- not a lot left at all !

How improve on this ? attack and reduce the 13% ? Hence the search / enquiry “Price Leverage “.

But “Price Leverage” has gone much further than achieving an equivalence of Odds = SP

To bring the this latter part to closure, I think - I need to rehash a selections sample at SP+13%, and use Price Leverage to bring the equilvent odds back down to SP

However, this is rather academic – since it presumes the selections were ALL STEAMERS……..and identified in sufficient time as being such.

Anyway - with the above brought to closure - then i will have 'cleared the space' - to 'let in' the full consequences of your post

thank you

[PeteB]

Fixador: No problem! I know what you mean about learning curve - so much still seems opaque to me, and I can remember a few months ago taking ages to work that lot out! Then you go back to a post or article you read a year ago and suddenly more of it makes sense...

I'm intrigued by the price leverage, and especially this example, because you swap your position. Are they suggesting that you always knew you wanted to lay this selection, but in spite of that backed it first because you thought you could gain leverage over the lay price you could get at the time? Or is it just a totally artificial example??

Incidentally, it seems to me with the price leverage, once the selection has steamed, you can effectively take whatever odds you want on either side.

If you want to end up backing, then you can lay off anything between 1p and £19.99, and you will be left with a back at certain effective odds - tends to infinity as your lay approaches £20. Your p
So if you lay £18, you are left with an effective back of £2 @ 14.3. If you lay £19, you are left with an effective back of back of £1 @ 26.3

Similarly, if you want to end up laying, then if you lay more than £38.46 you can choose your lay odds, with the odds tending to 1 the closer your stake is to £38.46.
So if you lay £39 you are left with an effective lay of £19 at 1.03, and if you lay £40 you are left with an effective lay of £20 at 1.1

(The maths for all this is the same if the price drifts - it's just that you can now lock in worse odds or a guaranteed loss)

So if you know the SP, it should be possible to work out how much to lay to get those odds (though I'm note sure how you know the SP at the point you are doing this??)

But if you are choosing your odds in this way, you can't choose your stake. You stake is determine by the difference in price between your initial back and the price you can now lay at. So you will be staking more (potentially many times more) on steamers than on selections that don't move or drift. This may blow the value of the selections, which may only work on level stakes - you would need a view on whether doing most of your laying (in terms of bet sizing) on steamers is a good idea.

It would be nice to know whether a steamer's real chances are better reflected by its initial price or its SP - there seems to be a view that the market gets more 'accurate' as the race approaches - is this true?? I've never looked at the data for that. It would probably be fairly easy if you had a history of racing post initial prices and SPs, and you could then see which had the stronger correlation to actual results...

[Fixador]

Hello Pete, ouch my brain hurts ....i have just figured out what to do if in the orginal example - the price drifted out - ie, greenup to minimise looses. And was wondering "As its drifted - Lay to green up the orginal Back bet, then Lay again in the hope it continues to drift - then close out /green up with a final Back bet "

Re my 80%WinReturn problem : Unfortunately - i need to be very careful here, as the selections i was orginally viewing ( results at SP ) , AND the example : which was from Betfair education - are different topics. I was welding them together - as a means to overcome / remove the SP+13% issue when Laying.

No one was directing me to start a Lay position, by Backing first - this came from the link below

....hope that makes sense.


The BF education link is here :
[url]
http://betting.betfair.com/education/th ... price.html
[/url]

That using "Price Leverage" - far exceeded my target of getting the equilvent Lay Odds back to the SP - is making my head spin

Pete - you are ofcourse correct - Cant know what the SP is, so targeting instead to evens 2.0 is probably more sensible

I've missed something............. back later

[Fixador]

Remebered now what i missed .....

The orginal 80%WinReturn selections did not mention steaming activity at all

Steaming entered the thread as a consequence of "Price Leverage" - as an additional complication ! from the BF education example

ie, i had exchanged a high certainty of the selection would Lose , but at Lay Odds too high,

for:-

Obtaining much lower Lay odds by using "Price Leverage" - BUT - the selection had to STEAM in ASWELL as LOSE

..... hummph !

still, with odds at evens - and the strikerate of 20% ( if backed ) - 80% Lay SR......

but again i confuse the issue - i can only get the odds to evens - IF the selection steams - but i dont know if steaming occurs ( of the orginal 80%WinReturn at SP - if backed)

i can go around this circle a few more times .

Break the impasse : back later - there is a way out - other than drop it in the bin !

Might be able to reply to yr 3rd paragraph & onwards - as well - ha ha !

[Fixador]

To summarise so far ( as i understand it ):-

(1) A selection has an 80% probabilty of Losing, but the Lay Odds are too high.

(2) Can respond by :-

(a) wait see if odds steam

(b) Set up the "Price Leverage" Back bet, and wait for odds to steam

(3) If the Odds steam: then have choice :-

(a) Complete 2(a) and Lay at the lower price

(b) Complete the "Price Leverage" - and Lay, and achieve equivlent odds of 2.0 ( evens ) or - as you pointed out all the way down to 1.03

(c) Go for FREE LAY - so offset the Back at 2(b) - and Lay at lower Odds with same stake

(d) Green Up the Back at 2(b) lock in profit before outcome known


so, 3(d) and 3(c) are the lowest risk

(4) If the odds drift, or dont move:-

(a) follow 2(a) - no bet

(b) Exit the back position of the "Price Leverage" - by a Green up Laybet to minimise losses


And as aside : Look for selections that steam, irrespective of outcome probability

Pete : I also had heard of some correlation between RP tissue price and SP - and have been recording them for a short while - havent had time to anylse as debugging mostly at mo. Hadnt thought to take it further - they bore a correlation to outcomes though! ha ha !

[PeteB]

Yep that's very neat and clear - thanks! (and thanks for the bf education link - I hadn't looked at that stuff before)

So I guess it comes down to where you find your edge... whether you are excellent at picking lays, or whether you are good at spotting potential steamers and drifters

BTW did you ever read Maria's Laying Thread (there is a link to it in a post by GeorgeUK from some time ago) - I learnt more from this than from anything else about how to turn a small edge into a large gain - fantastic!

[Shaun]

This is an interesting thread, would you have the username of the person that posted that thread i can't find it.

[Fixador]

It would be nice to know whether a steamer's real chances are better reflected by its initial price or its SP - there seems to be a view that the market gets more 'accurate' as the race approaches - is this true?? I've never looked at the data for that. It would probably be fairly easy if you had a history of racing post initial prices and SPs, and you could then see which had the stronger correlation to actual results

PeteB : somewhere in BF education, it says that RP checked BF prices just before the OFF, and conclude that those BF prices were a true reflection of the actual odds of winning. So - yes, market is accurate at OFF.

So - and thats a big long exhaling "So" ! Its the SP which is inaccurate ? !

Ok, a bold statement - everyone may disagree ! - just wondering out loud - is it the bookies overround that distorts the SP. Or, better stated as - the SP is distorted because of the bookies overound

I'm going to back out of this now - as , suddenly, the obvious has occurred to me. The SP has become irrelevent - since as you pointed out earlier - we dont know what it is - nor can we know. And now i am going to add - i dont want to attempt to know or rather, 2nd guess what it might be.

Far, far better, would be something we can know in advance - an f/c price

Should there be an approximate correlation - even over a limited range - of f/c to BF prices very close to the OFF , so, that if , on a future race- at a time prior to the OFF , the BF price is higher than the correlation with the f/c suggests - then, a statement of probability that the runner will steam in can be made - hence an opportunity to launch into "Price Leverage", or "Free Lay" or "Free Back" or "Green Up" exists



( and the paragraph above : thats one of those "swallow a camel easierly " statements - that is going to take a while to prove / disprove )

PeteB : reading back through this , i think you have already stated most this - but in far less words !

so, time to crank up the f/c scrapers

Raiding Massey's database as suddenly taken a twist for the worse - with the prices quoted at SP - and my suddenly not trusting SP anymore

[xraymitch]

Great discussion people many thanks for that, and I need some time to digest it

Shaun - I have come across Maria several times (in forums) over the years and I have a great respect for her.


You can find out more here from this link in a conversation with georgeuk

http://gruss-software.co.uk/forum/viewt ... ht=ratchet

if you can go follow the discussion on eba

I see Maria is now banned on punterslounge - which surprises me greatly.

[Fixador]

TWO bizarre things happened yesterday .

(1) The RP forecast price changed at least twice during the day:-

25 Aug : Cartmel 1530 “Rookery Lad” was 5/1 , then 11/2 , then 5/1

(2) When the f/c was 5/1, the jockey column , had got “DOUBTFUL” , and when the f/c was 11/2 , the the jockey was P J Brennan

I thought the f/c was fixed in value ?

AND today :-

at 0930 :-

BETTING FORECAST: 9/4 Needwood Lad, 5/1 Coleorton Choice, 6/1 Lisburn, 7/1 Firebet, 8/1 Harriet´s Girl, 10/1 Abbey Steps, 12/1 Fitzolini, 20/1 Madame Jourdain, Veronicas Boy, 25/1 Deadly Encounter, Pacific Bay, Toby Tyler, 50/1 Tagula Sunset


then from the USA ( at 11 am uk time - I teleported ! )

Betting Forecast: 3/1 Needwood Lad, 6/1 Lisburn (IRE), Fitzolini, 7/1 Coleorton Choice, 8/1 Harriet's Girl, 10/1 Firebet (IRE), Abbey Steps (IRE), 14/1 Veronicas Boy, 16/1 Madame Jourdain (IRE), 20/1 Pacific Bay (IRE), 25/1 Toby Tyler, Deadly Encounter (IRE), 33/1 Tagula Sunset (IRE)

Now back in the UK at 1400

BETTING FORECAST: 9/4 Needwood Lad, 5/1 Coleorton Choice, 6/1 Lisburn, 7/1 Firebet, 8/1 Harriet´s Girl, 10/1 Abbey Steps, 12/1 Fitzolini, 20/1 Madame Jourdain, Veronicas Boy, 25/1 Deadly Encounter, Pacific Bay, Toby Tyler, 50/1 Tagula Sunset


WHAT THE FLAMING HECK IS GOING ON !

Mr Quick ( but not amused )

[KevinTHFC]

I believe the RP site was being upgraded this morning, perhaps that had something to do with it.

I haven't noticed the F/C prices charge before, but then again I have never realy looked to see if they had!

Kevin

[KevinTHFC]

I believe the RP site was being upgraded this morning, perhaps that had something to do with it.

I haven't noticed the F/C prices change before, but then again I have never realy looked to see if they had!

Kevin