Trading What I see !?

[goat68]

Planning my next strategy, called "operation Order Flow" Python strategy! A set of classes to encompass various Webb based order flow aspects. Do you like my spin on web-based, quite like that one

I've been 'learning' order flow trading (in pre-off horse racing markets) for about 7 months now. I watched all Peter's videos on the subject repeatedly and have been manually trading all that time.

I lost money for the first 4 months but the loss decreased each month. I've have had a (small) profit for the last 3 months and I'm now at a position where my confidence is growing and I feel I can 'read' the markets much more clearly (though it's not easy). After 7 profitable days on the run I took a hit yesterday but that was down to poor trading by me - basically I dropped my guard psychologically and got mullered!!

My point is that, in my humble opinion, to learn order flow trading you'll have to sit for hour after hour watching - and ultimately trading - the markets. I'm not sure it can be automated - it is very nuanced and subtle. If aspects of it can be automated I'm a long way from seeing it but I'm becoming more comfortable with the manual side of it as time goes on.

Sounds like you're making progress, well done.
My plan is to come up with order flow models or behaviour and then bactest to verify it..
Im trying to get away from backtesting and hunting data for strategies, i don't think that works..

[goat68]

Of course I need market time to see this, so thinking of maybe setting up BA guardian to autoswitch markets and screen record to watch in the evening....

[Realrocknrolla]

Of course I need market time to see this, so thinking of maybe setting up BA guardian to autoswitch markets and screen record to watch in the evening....

Good idea. That works.

I did it a few times!

[agog]

Of course I need market time to see this, so thinking of maybe setting up BA guardian to autoswitch markets and screen record to watch in the evening....

Yes, I think recording the pre-off markets to study later is a great strategy for anybody trying to learn order flow. It's particularly helpful in reviewing your mistakes as sometimes it can be difficult to pinpoint where you've gone wrong in a fast-moving market.

[goat68]

Getting very frustrated at this again...
Spent the last few evenings working on the CrossOver point order flow type trade. So i've created a bot that detects previous race finishes, then within the following minute after a previous race finish for a market that is approaching start, determines momentum of the top runners, and if one of those is hovering around a crossover (3,4,6), and the other main runners (ones with significant volume%) are showing Backing momentum, then Lay that selection.
I have been trying numerous, hedging, take profit/partial profit, adding to trade,... akin to what Peter demonstrates in his videos...
And I have been testing over a month sample.
Result....nothing shows a profit......!

I do think I am trying to do the impossible, as thinking about it, if you watch Peter's videos, his decision making for example on where he places exit trades, is very subjective based on 20years of experience, I am not going to achieve that!!!
I have tried:
- Stop loss, hedge at start if open
- Stop loss, and fixed target profit, hedge at start if open
- Stop loss, and fixed 1/2 target profit, run the rest, hedging at start
- Stop loss, and add to trade at a target, hedging at start
- No stop loss, and add to trade at a target, hedging at start
As you can see, all basic rules really, nothing compared to a subjective 20years of experience..........!

Of course, all these exits are useless if there is no edge in my "entry" to start with...!

[goat68]

So pure-auto is too hard, and order-flow trading is really for the manual world, Hmmm

[napshnap]

goat68, few suggestions:

- keep in mind that different race types have different market behaviour. Hcap markets are less volatile that non-hcap or Irish markets. Don't put everything in the same bag.

- test crossovers of different moving averages with different lengths.

if there is no edge in my "entry" to start with

- try this basic aproach from financial trading: "If the price is above VWAP, it is a good intraday price to sell (back). If the price is below VWAP, it is a good intraday price to buy (lay)."

[adryan87]

When is the best time to enter on a trade pre-off?? Everytime i back,odd drift against me:)).It's a window time or something else ? How to anticipate an odd will drop?? No one want to say this things? )

[goat68]

goat68, few suggestions:

- keep in mind that different race types have different market behaviour. Hcap markets are less volatile that non-hcap or Irish markets. Don't put everything in the same bag.

- test crossovers of different moving averages with different lengths.

if there is no edge in my "entry" to start with

- try this basic aproach from financial trading: "If the price is above VWAP, it is a good intraday price to sell (back). If the price is below VWAP, it is a good intraday price to buy (lay)."

Thanks, i've tried VWAP, and it's variable, it works nicely for one week, then awful the next... it's another demonstration of how a manual trader would probably spot something different those bad weeks....?
Race type for Crossover points is an interesting question, I ask myself why race type would affect crossover behaviour? The fact that odds increments change affects both Hcap's and Mdn's...? Yes, one is more volatile than the other, so maybe things like leeway for noise and targets would differ...?

[goat68]

The key difference I see is with automation I have a strict unchanging set of rules, which can go horribly wrong if the markets change, whereas manual traders can react to the market changes they see occurring...

[adryan87]

Ok..this is flipping useless..at What the hell to look to anticipate a drop..i backed than odd drift..what the hell..how unlucky i am..

[Maturin]

goat68, few suggestions:


- try this basic aproach from financial trading: "If the price is above VWAP, it is a good intraday price to sell (back). If the price is below VWAP, it is a good intraday price to buy (lay)."

Do you mean for specifc race types here and/or certain times? Because from what I can see, prices often hold above/below VWAP for sustained periods.

[goat68]

goat68, few suggestions:


- try this basic aproach from financial trading: "If the price is above VWAP, it is a good intraday price to sell (back). If the price is below VWAP, it is a good intraday price to buy (lay)."

Do you mean for specifc race types here and/or certain times? Because from what I can see, prices often hold above/below VWAP for sustained periods.

yes agree

[adryan87]

So,it's a secret..

[Maturin]

Ok..this is flipping useless..at What the hell to look to anticipate a drop..i backed than odd drift..what the hell..how unlucky i am..

Welcome to trading!

Regarding your question about time to enter - you will see differnt periods of increased amounts of money coming in. Things can change rapidly when the money starts to flow in. Beware there are sometimes large spikes at 4 - 5 mins out (often seems to be 4:13 to me but not always). Then at 2 mins there seems to be a big influx of money, where prices can move quite dramatically in either direction. Observe some markets to see if you can pick up on it. As to predicting the direction...