Trading What I see !?

[goat68]

Liam is trying to say if you place your stop losses where it triggers more than the profit or the long term expectancy of the spread isn't positive, you won't make money in the long term.

spread = 1tick generally yes? eg.5.4 to 5.5 ?
what is "long term expectancy of the spread"?, spread is 1tick 5.4 to 5.5 ?!
maybe you mean something different with spread?

[goat68]

Liam is trying to say if you place your stop losses where it triggers more than the profit or the long term expectancy of the spread isn't positive, you won't make money in the long term.

if I can determine a market setup eg.WOM scalping, where the expectancy of 1tick profit is say 80%, and 3tick loss 20%, then over 100 bets I will make 20ticks.

I also find it a bit harsh that you're making comments about a strategy very similar to Dallas's ScalpingWOM, which actually uses a 4tick stop.

[goat68]

spread = "the spread is the gap between the bid and the ask prices of a security or asset"

[Realrocknrolla]

Liam is trying to say if you place your stop losses where it triggers more than the profit or the long term expectancy of the spread isn't positive, you won't make money in the long term.

When i read Peters posts, i read them in Raymond Reddington voice.

Both interesting in there own rights

[goat68]

Liam is trying to say if you place your stop losses where it triggers more than the profit or the long term expectancy of the spread isn't positive, you won't make money in the long term.

When i read Peters posts, i read them in Raymond Reddington voice.

Both interesting in there own rights

This demonstrates what a huge void there is between those that know and those that don't.
I guess this forum is never going to be a place of teaching, so why come down to the newbie level...
Anyway, as I said, I'm going to tweak my rules and hopefully improve.

[Jukebox]

This demonstrates what a huge void there is between those that know and those that don't.

You'll fair better if you ask questions rather than make statements

[goat68]

This demonstrates what a huge void there is between those that know and those that don't.

You'll fair better if you ask questions rather than make statements

I'll try .
Liam, Peter, please can you explain again in more basic terms, maybe an example?
Thanks

[jamesg46]

In order to be profitable you need to place positive ev bets, how is a stop order ever going to be positive ev long term?

It can't be that simple, so you're saying Guardian is crap as it provides tools that are by definition -ev ?
To me if timed in the right market, and you achieve a strike rate and risk/reward correctly you should be able to achieve +ev?

Why would an order that is crossing the spread on the opposite side of your initial (supposed +ve) be positive ev?

Not saying anything is crap but you can’t just apply order types / strategies without thinking about them objectively in terms of expected value.

There must be a translation issue from intelligent to stupid because I'm not sure I understand. From my stupid perspective it comes across that you never have a -ve position to close. That every position taken is opened with a +ev long term expectancy, which I believe you wouldn't open positions if it was a -ve expectancy, that would be a little silly, but there isn't a guarantee on a +ve expectancy because we are expecting.

Is this a way of saying to goat that he needs to find his long term +ve expectancy before he looks at crossing the spread with a stop loss order?

[jamesg46]

The way I understand the advise that's been given is that you don't know what your expectancy is yet, or even if you're entering at a point of +ve expectancy at the moment you enter the current spread.

So why would you put a stop loss order into the market? Wouldn't it be better to enter at a chosen point because of your xyz idea and then prove both the +ve & the -ve expectancy. If you use a stop order while trying to prove your +ve expectancy then whats to say you're not prematurely stopping out of positive positions?you simply do not know because the stop orders that are triggered without first knowing your +ve expectancy are distorting whats possible long term in exchange for short term gain, this will never be profitable.

[LinusP]

Sure, let's start at the beginning, to make a profit you need to get value (positive expected value) it comes down to simple maths that without it you won't be profitable. This applies to betters, gamblers, traders, scalpers, swing traders... what ever you want to call yourself it is mathematically impossible to be profitable long term without getting value.

A simple example is a fair coin that should be priced at 50:50, on befair this might look like this:

Runner | back | lay
Heads | 1.99 | 2.02
Evens | 1.99 | 2.02

You model this market and calculate that Heads at 2.02 is giving you a positive 0.5% edge, so you place a bet which sits in the queue and eventually gets matched. At this point it would be pretty stupid to offset at 2.00 as that is the correct price (0% ev) so I imagine that instead you offset at 1.99 (note that I think offset bets are also pretty stupid and you should instead place all bets off modelling/triggers)

Suddenly the market moves, this could be due to the mad bomber being released or someone with more information on the fairness of said coin, none of this changes your initial bet but the market now looks like this:

Runner | back | lay
Heads | 2.08 | 2.10
Evens | 1.89 | 2.02

Obviously backing heads at 2.08 is an obvious +ev bet based on your modelling and you would want to get fully stuck into taking this however instead your naive stop loss is going to be hit and you will instead lay 2.10 (-2.4% edge) resulting in a red screen and an overall -1.9% ev.


You have to treat each bet individually regardless of what you are doing, simply calling it trading does not hide the fact that you need to take value long term to be profitable. In the real world it is a case of recording all your bets / triggers / market state and then analysing the results to find where the value is and where it isn't, however it is very easy to see that simple stop losses are never going to be +ev.

[goat68]

Sure, let's start at the beginning, to make a profit you need to get value (positive expected value) it comes down to simple maths that without it you won't be profitable. This applies to betters, gamblers, traders, scalpers, swing traders... what ever you want to call yourself it is mathematically impossible to be profitable long term without getting value.

A simple example is a fair coin that should be priced at 50:50, on befair this might look like this:

Runner | back | lay
Heads | 1.99 | 2.02
Evens | 1.99 | 2.02

You model this market and calculate that Heads at 2.02 is giving you a positive 0.5% edge, so you place a bet which sits in the queue and eventually gets matched. At this point it would be pretty stupid to offset at 2.00 as that is the correct price (0% ev) so I imagine that instead you offset at 1.99 (note that I think offset bets are also pretty stupid and you should instead place all bets off modelling/triggers)

Suddenly the market moves, this could be due to the mad bomber being released or someone with more information on the fairness of said coin, none of this changes your initial bet but the market now looks like this:

Runner | back | lay
Heads | 2.08 | 2.10
Evens | 1.89 | 2.02

Obviously backing heads at 2.08 is an obvious +ev bet based on your modelling and you would want to get fully stuck into taking this however instead your naive stop loss is going to be hit and you will instead lay 2.10 (-2.4% edge) resulting in a red screen and an overall -1.9% ev.


You have to treat each bet individually regardless of what you are doing, simply calling it trading does not hide the fact that you need to take value long term to be profitable. In the real world it is a case of recording all your bets / triggers / market state and then analysing the results to find where the value is and where it isn't, however it is very easy to see that simple stop losses are never going to be +ev.

Hi Linus, thank you very much for writing this up, it does help.
So I understand what you're saying now, when you say crossing over the spread, you're saying the stop distance goes from the back bet price 'over' the one tick spread. I would agree having a one tick stop on an entry like this would be stupid.
However, in my example I am using a 3 tick stop and entering at the best reverse price if matched. So in your example I am a queued back bet at 2.10, and a stop at 2.16. My premise being I've identified a slight directional bias at this time such that my 1tick profit (2.08) occurs over 3times more often than the 3tick loss.

I do understand what you're saying James as well, in that I don't really know this is a +ev model, so why have a stop,ie.let it run till post time, but then that might not be +ev either, or preferably develop a closing trade model that makes it +ev, equally working that out is hard but better.

Thanks again, I got there in the end (I think!)

[goat68]

Sure, let's start at the beginning, to make a profit you need to get value (positive expected value) it comes down to simple maths that without it you won't be profitable. This applies to betters, gamblers, traders, scalpers, swing traders... what ever you want to call yourself it is mathematically impossible to be profitable long term without getting value.

A simple example is a fair coin that should be priced at 50:50, on befair this might look like this:

Runner | back | lay
Heads | 1.99 | 2.02
Evens | 1.99 | 2.02

You model this market and calculate that Heads at 2.02 is giving you a positive 0.5% edge, so you place a bet which sits in the queue and eventually gets matched. At this point it would be pretty stupid to offset at 2.00 as that is the correct price (0% ev) so I imagine that instead you offset at 1.99 (note that I think offset bets are also pretty stupid and you should instead place all bets off modelling/triggers)

Suddenly the market moves, this could be due to the mad bomber being released or someone with more information on the fairness of said coin, none of this changes your initial bet but the market now looks like this:

Runner | back | lay
Heads | 2.08 | 2.10
Evens | 1.89 | 2.02

Obviously backing heads at 2.08 is an obvious +ev bet based on your modelling and you would want to get fully stuck into taking this however instead your naive stop loss is going to be hit and you will instead lay 2.10 (-2.4% edge) resulting in a red screen and an overall -1.9% ev.


You have to treat each bet individually regardless of what you are doing, simply calling it trading does not hide the fact that you need to take value long term to be profitable. In the real world it is a case of recording all your bets / triggers / market state and then analysing the results to find where the value is and where it isn't, however it is very easy to see that simple stop losses are never going to be +ev.

Just one other thing though, with regards value trading, I questioned Peter over whether his pre race horses was in any way value trading and he said no. As with horses there is no correct modelled price, it can keep coming in even if the horse will never win... A fair coin has physical characteristics in the real world which means evens price is the expected model.

[jamesg46]

And here comes the crossover between discretionary & quant

[goat68]

I would add on a more positive note, I am quite enjoying Guardian Automation, which as per JG's guide is important...
Manual trading was just frustrating me!

[Trader Pat]

I guess this forum is never going to be a place of teaching, so why come down to the newbie level...

I mean this in the nicest possible way Goat and its not meant as a criticism but I think you need to slightly readjust your attitude if you're going to make a real go of this.

When you're manual trading you seem to think there's some kind of conspiracy and that people are keeping stuff from you (they aren't) but automated trading is a completely different story. People work for months even years on their automation and their not going to share their secrets with anyone and nor should they.

My advice to you would be to stick to manual trading until you get a better understanding of the market. If you think manual trading is frustrating you'll probably find automated trading a 1000 times more frustrating.

As for your stop loss query I think 3 ticks is too tight unless you're trading something like the Grand National, 5 ticks would be more sensible but even then you wont make a profit long term unless you're very selective with your entry criteria. And getting your entry criteria right comes down to a better understanding of the market. So like I said I'd stick to manual trading, if you're getting frustrated its probably down to overstaking, reduce your stakes and if you're still getting frustrated then you may have to admit that trading isn't for you.