When Trading Meets DividendWarrior

[Shiny Things]

Well, if u are interested, u can try looking into this idea of the strange attractor in chaos mathematics, it's really interesting, a chaotic system has a strange attractor where states around it is always attracted and inevitably attracted to it, and that may mean, up to a point the markets are random and noisy in the short terms but inevitably goes somewhere in the long term

Hang on, hang on. I'm getting a bit out of my depth here, because this is getting into quant-land, but I wanted to pull you up on this. You're just trying to generate some useful out-of-sample data to test a trading system. Saying "let's use chaos mathematics!" seems like you'd be adding a lot of unnecessary complexity.

If you're trying to say "short-term moves are random but long-term moves are trended", couldn't you save a hell of a lot of time and model complexity by using a garden-variety random walk with a nonzero drift term?

Also help me out, I'm a little bit confused by this:

average the position while keeping the put ratio 2 to 1 to spot so jta a long straddle and then if possible sell 1 put 1 call and become a box spread and just let it expire worthless

I've missed something blindingly obvious here. A box spread is two synthetic forwards at different strikes. If you start off long 2 puts at the same strike, and then you sell a put and a call, you're not going to end up with a box spread no matter what strikes you use.

Can you walk me through this?

 

[CookieMonsta88]

Hang on, hang on. I'm getting a bit out of my depth here, because this is getting into quant-land, but I wanted to pull you up on this. You're just trying to generate some useful out-of-sample data to test a trading system. Saying "let's use chaos mathematics!" seems like you'd be adding a lot of unnecessary complexity.

If you're trying to say "short-term moves are random but long-term moves are trended", couldn't you save a hell of a lot of time and model complexity by using a garden-variety random walk with a nonzero drift term?

Also help me out, I'm a little bit confused by this:



I've missed something blindingly obvious here. A box spread is two synthetic forwards at different strikes. If you start off long 2 puts at the same strike, and then you sell a put and a call, you're not going to end up with a box spread no matter what strikes you use.

Can you walk me through this?

Well, u could, but I would model it after a variation of Elliot wave theory to model the nature of impulse waves made by strong hands and corrective waves and the fractal nature of the markets and then counter impulse waves must also be modelled for so I don't think Gaussian models which are linear handles it, its a non linear branch of modeling, I read it from the black swan by nassim taleb.

Well complexity in exchange for accuracy is desirable here since we are dealing with money, I'd go the extra mile just for it

Try modeling the market's price action intuitively with 1 manipulator and 5 other weak holders, I think u will get my point

And as for the 2puts, there 1 spot too, so its a long straddle, long 1 straddle short 1 straddle becomes a box spread no? 1put +1 spot = 1 call, 1 long call + 1 long put +1 short call +1 short put = box spread

But its carried out this way as a failsafe against counter trend moves, as u start off with puts of 10x bigger than ur current spot size, then u average into ur puts to become 2 to 1 ratio

 

[wahkao3]

the quanty delta de carlo chaos math ethics is actually a derivative of the integral of the sine wave of the fundamental frequency.

i read till gong gong liao

 

[Nanonited]

Well, u could, but I would model it after a variation of Elliot wave theory to model the nature of impulse waves made by strong hands and corrective waves and the fractal nature of the markets and then counter impulse waves must also be modelled for so I don't think Gaussian models which are linear handles it, its a non linear branch of modeling, I read it from the black swan by nassim taleb.

Well complexity in exchange for accuracy is desirable here since we are dealing with money, I'd go the extra mile just for it

Try modeling the market's price action intuitively with 1 manipulator and 5 other weak holders, I think u will get my point

And as for the 2puts, there 1 spot too, so its a long straddle, long 1 straddle short 1 straddle becomes a box spread no? 1put +1 spot = 1 call, 1 long call + 1 long put +1 short call +1 short put = box spread

But its carried out this way as a failsafe against counter trend moves, as u start off with puts of 10x bigger than ur current spot size, then u average into ur puts to become 2 to 1 ratio

hey cookie!! your inbox is full please clear if possible thanks!!

 

[Shiny Things]

Well, u could, but I would model it after a variation of Elliot wave theory to model the nature of impulse waves made by strong hands and corrective waves and the fractal nature of the markets and then counter impulse waves must also be modelled for so I don't think Gaussian models which are linear handles it, its a non linear branch of modeling, I read it from the black swan by nassim taleb.

I think we'll have to agree to disagree about this one, but don't forget: you're just building a market model to generate testing data for strategies here. You don't want to be chewing up huge amounts of time building a wonderfully elaborate market model for testing trading strategies; you want to be spending that time building the trading strategies.

And as for the 2puts, there 1 spot too, so its a long straddle, long 1 straddle short 1 straddle becomes a box spread no? 1put +1 spot = 1 call, 1 long call + 1 long put +1 short call +1 short put = box spread

No.

A box spread is two synthetic forwards: it's "long call + short put at strike x, and long put + short call at strike y≠x".

If you're looking at it in a delta-hedged way (which I assume you are, because it looks like you've got a grasp on put-call parity), a box spread is basically a flat position, but not quite - it's a synthetic loan, rather than a dead-flat position.

If you start off long 2 puts at strike x, then you sell a put and a call at strike y≠x, that's not a box spread - it's a risk reversal (or a collar; different names for the same thing).

If you start off long 2 puts at strike x, then you sell a put and a call at strike x, then you're flat. You're long a put and short a put (those net out); and long a put and short a call (that's a synthetic forward); and presumably you're long the full amount of delta against the synthetic forward. So you're flat, but it's not a box spread.

 

[CookieMonsta88]

hey cookie!! your inbox is full please clear if possible thanks!!

Eh, sry, try send i just cleared it

 

[CookieMonsta88]

I think we'll have to agree to disagree about this one, but don't forget: you're just building a market model to generate testing data for strategies here. You don't want to be chewing up huge amounts of time building a wonderfully elaborate market model for testing trading strategies; you want to be spending that time building the trading strategies.



No.

A box spread is two synthetic forwards: it's "long call + short put at strike x, and long put + short call at strike y≠x".

If you're looking at it in a delta-hedged way (which I assume you are, because it looks like you've got a grasp on put-call parity), a box spread is basically a flat position, but not quite - it's a synthetic loan, rather than a dead-flat position.

If you start off long 2 puts at strike x, then you sell a put and a call at strike y≠x, that's not a box spread - it's a risk reversal (or a collar; different names for the same thing).

If you start off long 2 puts at strike x, then you sell a put and a call at strike x, then you're flat. You're long a put and short a put (those net out); and long a put and short a call (that's a synthetic forward); and presumably you're long the full amount of delta against the synthetic forward. So you're flat, but it's not a box spread.

Actually, ur perfectly correct on the model thing and I do agree with u on that, so I think the best way is just take use a demo account from a broker and forward test it, save the trouble as u mentioned above

Yea, I need to model this more, but the rough intuition is this, so I might use reversal or condors or butterflies, but I need to study more on this method, only figured this a few days ago, its still very sketchy and yea flat is a more appropriat e term here

 

[Shiny Things]

Actually, ur perfectly correct on the model thing and I do agree with u on that, so I think the best way is just take use a demo account from a broker and forward test it, save the trouble as u mentioned above

Yea, I need to model this more, but the rough intuition is this, so I might use reversal or condors or butterflies, but I need to study more on this method, only figured this a few days ago, its still very sketchy and yea flat is a more appropriat e term here

Cool, no worries. All you're doing here in your example is "buying 2 options to protect when things are volatile" and then "selling them back later to get flat"; I think you're overthinking it a bit. (And you're also buying vega when vol's high and selling it when it's low, so that's going to be an expensive hedging strategy.)

Just politely mate, I really think you might be trying to run before you can walk when it comes to options. Learn a bit more options theory; pick yourself up a demo account and do some paper-trading; and feel free to ask on here if you've got any more questions.

 

[CookieMonsta88]

Cool, no worries. All you're doing here in your example is "buying 2 options to protect when things are volatile" and then "selling them back later to get flat"; I think you're overthinking it a bit. (And you're also buying vega when vol's high and selling it when it's low, so that's going to be an expensive hedging strategy.)

Just politely mate, I really think you might be trying to run before you can walk when it comes to options. Learn a bit more options theory; pick yourself up a demo account and do some paper-trading; and feel free to ask on here if you've got any more questions.

Yes, what u pointed out is the problem I'm having now, I have an option to just let it ride out without selling the option, just take as a lottery ticket, so yea, still figuring

Thx mate, been reading a book by Charles cottle, great options book

 

[Knight_Rider]

Actually, ur perfectly correct on the model thing and I do agree with u on that, so I think the best way is just take use a demo account from a broker and forward test it, save the trouble as u mentioned above

Yea, I need to model this more, but the rough intuition is this, so I might use reversal or condors or butterflies, but I need to study more on this method, only figured this a few days ago, its still very sketchy and yea flat is a more appropriat e term here

Results of a back test or forward test from a demo account account will not yield the actual results from a actual account. Even cents and normal account chart is different. You could test it on a $200 account and see how it goes.

 

[CookieMonsta88]

Results of a back test or forward test from a demo account account will not yield the actual results from a actual account. Even cents and normal account chart is different. You could test it on a $200 account and see how it goes.

As a matter of fact I am very interested in trying it out, but I am stuck with regards to a reliable broker which offers vanilla options for forex and spot forex. Any recommendations?

 

[Knight_Rider]

As a matter of fact I am very interested in trying it out, but I am stuck with regards to a reliable broker which offers vanilla options for forex and spot forex. Any recommendations?

For forex a couple of good ones like capitaltrustmarket or dukascopy. Local UOB is also into forex. Spot sorry no experience.

 

[CookieMonsta88]

For forex a couple of good ones like capitaltrustmarket or dukascopy. Local UOB is also into forex. Spot sorry no experience.

ic, i can check out with UOB, im using oanda now for spot, i think ikonfx has options and forex together, but as for how reliable and regulated they are, i have no idea, hence i dare not go sign up with them. thx for the info

 

[Knight_Rider]

IG has the most instruments but the worst spread. So use the right broker for the right strategy. Sports arbitrage used to be the most profitable using software to find the best odds. As long as there is a difference there is money to be made. I think it's more difficult now.

 

[Shiny Things]

As a matter of fact I am very interested in trying it out, but I am stuck with regards to a reliable broker which offers vanilla options for forex and spot forex. Any recommendations?

There aren't a lot of brokers that offer FX options to retail. You've got a couple of options:
1) Saxo: they're surprisingly good in this little niche - wide range of currencies, competitive pricing - but they're way too expensive everywhere else;
2) IkonFX, IG, all the other CFD brokers and spread-betting shops, etc etc etc - no. Just no. (Ikon touts a minimum 10-pip spread in EURUSD options, which is hilarious - that's about twice as wide as it should be.)
3) CME. This might be the best way to go: the CME's FX options market is deep and liquid, but the liquidity's concentrated in the majors. (The only quirk is that the options are on futures, not on spot, so you're effectively trading options on the outright forward; but the end result is still the same.)

 

[CookieMonsta88]

There aren't a lot of brokers that offer FX options to retail. You've got a couple of options:
1) Saxo: they're surprisingly good in this little niche - wide range of currencies, competitive pricing - but they're way too expensive everywhere else;
2) IkonFX, IG, all the other CFD brokers and spread-betting shops, etc etc etc - no. Just no. (Ikon touts a minimum 10-pip spread in EURUSD options, which is hilarious - that's about twice as wide as it should be.)
3) CME. This might be the best way to go: the CME's FX options market is deep and liquid, but the liquidity's concentrated in the majors. (The only quirk is that the options are on futures, not on spot, so you're effectively trading options on the outright forward; but the end result is still the same.)

actually im curious, since fx i have to go through a market maker and im always being shaded, this method should be stronger in exchange based markets with no middle man, since i can make markets and collect the spread edge and have a built in edge.

 

[CookieMonsta88]

IG has the most instruments but the worst spread. So use the right broker for the right strategy. Sports arbitrage used to be the most profitable using software to find the best odds. As long as there is a difference there is money to be made. I think it's more difficult now.

ic, if spreads are wider, i think need to exploit the larger swings if not the spreads can kill the account.

 

[Shiny Things]

actually im curious, since fx i have to go through a market maker and im always being shaded,

You're probably not. Unless you're doing clips of $5mio or more, it's not worth their time to figure out how to shade the price against you, and I don't think most shops bother with customer-level price shading. (There'd be no point in having one anyway - if you shaded a customer who trades direct, but you don't shade one who trades through, say, IBKR or a white-labeler, then you'd just trade through the white-labeler and get the unshaded price every time.)

this method should be stronger in exchange based markets with no middle man, since i can make markets and collect the spread edge and have a built in edge.

Yeah, but don't get any ideas about doing this regularly. Market-making in very liquid stuff like index futures or FX is not for retail investors. You can work bids and offers to capture the spread, sure (Interactive's FX platform lets you do this, and you can interact with orders from other IBKR customers) but there are high-frequency firms that spend eight- or nine-figure amounts building dedicated market-making systems. You can't compete with them full-time.

 

[CookieMonsta88]

You're probably not. Unless you're doing clips of $5mio or more, it's not worth their time to figure out how to shade the price against you, and I don't think most shops bother with customer-level price shading. (There'd be no point in having one anyway - if you shaded a customer who trades direct, but you don't shade one who trades through, say, IBKR or a white-labeler, then you'd just trade through the white-labeler and get the unshaded price every time.)



Yeah, but don't get any ideas about doing this regularly. Market-making in very liquid stuff like index futures or FX is not for retail investors. You can work bids and offers to capture the spread, sure (Interactive's FX platform lets you do this, and you can interact with orders from other IBKR customers) but there are high-frequency firms that spend eight- or nine-figure amounts building dedicated market-making systems. You can't compete with them full-time.

ic, though sometimes i find shading quite informative also, they always shade people on the wrong side of the trend.

well, cool, though i wasn't thinking about beating algo houses, more of trying to exploit the gamma curves, so i think the average trade might last a month, so not in direct competition with them.

 

[Shiny Things]

ic, though sometimes i find shading quite informative also, they always shade people on the wrong side of the trend.

Hmm? It's not about who's on which side of the trend; market-making algorithms skew their price depending on whether the algo itself is long or short. If the algo gets long, it'll move its price lower relative to the market at large so that it can get paid; vice versa if it gets short.

well, cool, though i wasn't thinking about beating algo houses, more of trying to exploit the gamma curves, so i think the average trade might last a month, so not in direct competition with them.

If your average trade lasts a month, then trying to chisel an extra tick or two out of the spread (and risking not getting filled) isn't a good idea. Don't worry about fancy market-making stuff. Just pay the offer.