Composite Hedge Arbitrage

This is a system I am currently exploring and working to perfect. First let me
explain the two goals:

1) protection of capital.
2) little to no risk returns due to swap margin.

To introduce the concept, lets consider a fictitious case where a bank will lend you $10,000 at 6% interest. Then you put your money in another bank, possibly in a CD or money market account paying 8% interest. You collect the 2% difference. Small but relatively risk free. I believe such a system is also possible in forex with swaps.

Since I started working on this, I searched the web and forums to see if anyone else is doing it, because it seems so obvious that I'm sure someone is. But I've not seen it yet. One system uses two brokers, one that pays a swap, and one that doesn't, but then you are constantly transferring money from one to another. A bit of a pain if you ask me.

Most traders are familiar with what I'd call a straight hedge, in that if you buy say, one lot of USDJPY, you also sell one lot of USDJPY so that your equity always remains the same. Obviously this is not a good strategy because the transaction costs (spread) and swap difference will lose you money. Brokers always make sure there is a gap in the swap between a long and short position. BUT thats where composites come in. Let me explain.

A composite is a two position trade, that hedges another single position. Back with the USDJPY position, say we are long. A composite to hedge the short side might be CADJPY and USDCAD. The CADs basically cancel out and you are really trading the USDJPY again. If you are long USDJPY, you are buying the USD, selling the JPY, so to hedge it you would be short the CADJPY (or buying the JPY, thereby offsetting your selling yen on USDJPY) and also short the USDCAD (selling the USD you bought with USDJPY).

Now you may ask WHY would you do this when you just made three trades, extra transaction costs and extra headache? Because the COMPOSITE pays more swap then the straight hedge of USDJPY, or charges less depending on the situation.

Maybe an example will help:
right now I am testing this with GBPAUD.
Short, it pays .81 per lot.
A straight hedge of going long GPDAUD would cost me -.86 or a total of -.05 ..no good.
Long GBPUSD pays me +.01 per lot (buying GBP)
Short AUDUSD pays -.54 per lot (selling AUD)
Composite total is -.53

.81 - .53 = .28 swap per day
Equity hedged and protected.

Now, caveats that need to be addressed:

1) three orders need to be executed basically simultaneously, so best to do by code with a script, or at VERY slow market times.
2) some of the crosses and minors have a liquidity difference so the hedging
doesn't always work out perfect.

I do plan, if this seems worth the time, to code a script, and also an indicator that will take a comma delimited file of swap rates and identify the most profitable composite/cross pairing to trade. First I'm going to test it for a while first. But I wanted to start a thread here and see if others have tried this or even heard of it, or can identify a hole in my logic.

Welcome comments or suggestions!

If it seems too good to be true, it usually is.

Nope. You would need more than 1 lot of aud/usd to be completely hedged. More like 1.7 lots and that would mean your paying more like .91 on that which means the entire thing is negative interest. If you trade your example they way you laid it out you will have some exposure to aud/usd moves.

I guess lots of things sound good in theory but don't work out so great in practice. Thats why you test out these things, and its worth the time because thats how new things are discovered all the time in many disciplines, math and science to finance.

Your points are valid, and I noticed in my testing that the composite does not completely hedge the main cross, as you pointed out. But it doesnt seem to be able to be adjusted by changing the lot sizes. For instance, the hedge to composite in just one day yesterday, had a range of 2 to 40 pip difference which would be unacceptable at large lot sizes needed to profit from the swap difference.

I think that has more to do with the liquidity levels, but either way it doesn't seem to be a worthwhile system to pursue further, IMO. Guess just have to stick with mix of fundamentals and technicals

Nice idea, here is an abitrage system in the UK using credit cards that does work.

It works like this:

1) Consumer opens credit card with 0% APR introductory offer

2) Consumer transfers balance from credit card into savings account

3) once introductory offer expires consumer pays off credit card and keeps the interest earnt.

4) consumer repeats the process with a new credit card.

The system is good because you don't even need any initial outlay, just a good credit rating.