ASX Stock Pairs Trade Journal

[nizar]

I agree with MRC, doesn't matter how high % your win rate is, you don't want 1-2 losses to wipe out all your gains.

You would need a stop maybe 4 Sd away from the mean which would like never ever get hit, but when it does, you'd be very happy you had it there.

Worth testing at least

Or have you already tested it and concluded that no stop works best?

 

[Pairs Trader]

What about the unexpected though, as in the case of Aleck, would you simply hold and assume eventually, they will come back to a mean, even though the fundamentals of the pair have obviously altered?

Personally, my account leveraged many times over would not be a problem, if I was comfortable with my directional trading and employed tight stops on individual trades. It's how futures scalping works.

But I like the general premise of your trading, just not the no stop part, correlations come and go and if it's thought this may be the case due to a change in the fundamentals of the correlated pair, I can't see why averaging further offside would be wise, in comparison with simply cutting the loss there and then.

Yes I can appreciate your view, its not wrong its just that I don't like using stops for various reasons, Im fine if one of my positions moves against me 50% or more it won't take me out of the game, as Ive said successful trading is all about making heaps of small trades with an edge, a pair will always return to the mean because we use a rolling mean not a static mean. If I were to employ a stop loss it would be time-based not price-based.

 

[Pairs Trader]

I agree with MRC, doesn't matter how high % your win rate is, you don't want 1-2 losses to wipe out all your gains.

You would need a stop maybe 4 Sd away from the mean which would like never ever get hit, but when it does, you'd be very happy you had it there.

Worth testing at least

Or have you already tested it and concluded that no stop works best?

In most trading systems stops generally decrease performance plus the added slippage and spread costs. If a pair were at 4stdev from its mean without significant reason I would be inclined to add more to the position not take it off.

 

[fapturbo]

In most trading systems stops generally decrease performance plus the added slippage and spread costs. If a pair were at 4stdev from its mean without significant reason I would be inclined to add more to the position not take it off.

How do you measure a Standard Deviation from it's mean?

 

[MRC & Co]

Interesting Pairs, it's definately thinking outside the box. Will be great to watch this thread play out with the various pairs trades.

Perhaps you should include the lot sizes, so then when you average down, it's easier to follow your average price?

 

[Largesse]

I'm a big fan of this method of trading.
Great for diversifying my trading methods as I primarily trade as a scalper intra-day.
The longer time frames mean I can do my analysis after-hours on EOD data and set my trades for the next morning

 

[Nick Radge]

Pairs,
Do you go like for like or ratio them up?

And would you cover this somewhere near 0.92?

 

[oby_2]

Hi All,

Very interesting conversation.

Pairs Trader - Do you invest $ for $ (eg. $10K in RIO and $10K in BHP) or do you use a Beta to calculate so a given % change in one should result in the same % change in the other?

Regards,
oby

 

[Pairs Trader]

How do you measure a Standard Deviation from it's mean?

I don't measure anything, the program does it for me.

 

[Pairs Trader]

Interesting Pairs, it's definately thinking outside the box. Will be great to watch this thread play out with the various pairs trades.

Perhaps you should include the lot sizes, so then when you average down, it's easier to follow your average price?

thanks MRC, yes I will include lot sizes if I average down on a trade so you can easily see the average, I don't avg down on every trade only if the pair really gets out of whack without a warranted cause.

 

[fapturbo]

I don't measure anything, the program does it for me.

Ok fine.

Do you know how the program measures it?

 

[Pairs Trader]

Pairs,
Do you go like for like or ratio them up?

And would you cover this somewhere near 0.92?

Ratio, at the moment a exit signal would be triggered around 0.9, however we use a rolling mean not a static one, so the exit level changes daily.

 

[Pairs Trader]

Ok fine.

Do you know how the program measures it?

Its the stdev of the ratio. I have it configured to measure the stdev of last 14days.

 

[Nick Radge]

exit signal would be triggered around 0.9

Wasn't far out! and yes I see why it changes.

 

[Pairs Trader]

Hi All,

Very interesting conversation.

Pairs Trader - Do you invest $ for $ (eg. $10K in RIO and $10K in BHP) or do you use a Beta to calculate so a given % change in one should result in the same % change in the other?

Regards,
oby

Yes I trade $ neutral, Im not sure if beta neutral improves performance as beta is constantly changing, I prefer to keep it simple dollar for dollar.

 

[Largesse]

not doing too badly on your trade Pairs Trader, up a nice 2.3% based on today's close.

 

[beerwm]

How about this example;

ABC Drug co. - 0.11c
XYZ Drug co. - 0.09c

90% + correlation, diverges/ etc - entry signal triggered

so you short-
ABC Drug co. - 0.11c
long-
XYZ Drug co. 0.09c

ABC Drug co - 0.11c - develops miracle drug, - surges to $11.
XYZ Drug co - 0.09c - stays put

so you put 1k, into both companies
your loss = 100,000 - [2k initial investment], = 98,000 loss.

-yet you 'cant' exit because the stocks havent returned to a correlation, but kept on diverging.

I'm sure your system works, but have you planned/experienced a situation like I suggested?
Reliance on stocks to continue a correlation after a divergence cant be guaranteed.... can it? especially when shorting.

sorry if i misunderstood your system, im happy to be corrected

 

[Largesse]

i'm sure one of his rules is: stay away from penny dreadfuls

to be honest i think your example is stupid, it's quite easy to throw up some extra-ordinary/ridiculous hypothetical that will make any process/method/set of rules invalid.

 

[beerwm]

there are plenty of shares that make 100x gains, even a 10x gain could cripple.

The reason why my example makes Pair's system invalid is because there is no cap to what he can lose. [like others have stated]

If the divergence continues.... you sit.

Judging by the RIO/BHP trade, it seems Pair doesnt consider fundamentals a crucial part of the system.

I dont think saying, that 2 stocks can diverge to a citical extent is "extra-ordinary/ridiculous hypothetical"

Anyway its just something to consider.

 

[Pairs Trader]

ABC Drug co - 0.11c - develops miracle drug, - surges to $11.
XYZ Drug co - 0.09c - stays put

so you put 1k, into both companies
your loss = 100,000 - [2k initial investment], = 98,000 loss.

-yet you 'cant' exit because the stocks havent returned to a correlation, but kept on diverging.

This would never happen to me because I only trade large caps, Ive never seen a large cap surge 2x overnight let alone 100x overnight, that is quite a ridiculous claim actually, you says this happens often, I would like to see some recent examples please. Plus whats the difference if you were just short the stock without a hedge, the same thing would happen, so shorting any stock exposes you to this risk, at least with pair trading you have removed market & sector risk leaving you only stock specific risk most of which can be avoided, don't trade into earnings, announcements, etc......yeah sure every now and then your going to find yourself in a ugly trade, that happens with any system and its part n parcel of trading.