Three way Hedge | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Money management in trading - the stupid way and the right way 22 replies

Hard Way, Easy Way 115 replies

Easiest way to hedge with silver/gold against EUR dropping? 0 replies

2 Way VS 1 Way Analysis 9 replies

Hedge strategy (not a hedge fund) 13 replies

Attachments: Three way Hedge
Exit Attachments

Three way Hedge

  • #1
  • First Post: Dec 19, 2006 5:26am Dec 19, 2006 5:26am
Searching about Hedging strategies, came accross a techinique which seems interesting (i.e. interest positive). The strategy is to to Buy 1 pair GBP/JPY, sell 1 pair USD/JPY and sell 1 pair GBP/USD. In this triangular hedge, the off balance in the units of Yens would be the volatility,which could be encashed on a positive swing, while all the time accruing imterest@ 6 USD per day per standard lot. Could some experienced trader throw light on this and also please let us know what is the maximum pips negative swing they have experienced or should be expected. Thanks
  • #2
  • Dec 19, 2006 6:27am Dec 19, 2006 6:27am
1 lot buy GBP/JPY = 1 lot buy GBP/USD + 1 lot buy USD/JPY

This is a typical misbelief by traders who do not think / check it more carefully.

Forget about the $6 / day interest, the fomula itself can be a big killer.
凹凸 only
  • #3
  • Dec 19, 2006 6:38am Dec 19, 2006 6:38am
Ok!, If you believe so, but could you pls elaborate how?. What is the catch? The units the behaviour, or what? Thanks
  • #4
  • Dec 19, 2006 6:52am Dec 19, 2006 6:52am
Yes, at least it's not 1 / 1 / 1 otherwise you are only partly edged as 1 lot with the different base currencies doesn't have the same value. So you have to do some math using the ratios between the different currencies.
  • #5
  • Dec 19, 2006 6:57am Dec 19, 2006 6:57am
It will be a big topic to elaborate;-)

In short, the factor "1" of GBP/JPY lot(s) in the formula is a variable, which means not a constant "one" or whatever.

At a specific time, you can not decide / foresee the future GBP/JPY movement would be influenced by USD/JPY or GBP/USD movements more. And sometimes they add up, and sometimes they deduct.

So, that is why crosses are served. Otherwise, you trade majors only.

Hope it clarifies your question a little.
凹凸 only
  • #6
  • Dec 19, 2006 1:38pm Dec 19, 2006 1:38pm
As I had pre-clarified in my first post that this is not a perfect hedge and thus the volatility in it is what makes it tempting for a trade. I wanted to know what could be the max swing experienced by anyone trading this hedge. Thanks
  • #7
  • Feb 23, 2007 6:52pm Feb 23, 2007 6:52pm
Quoting Samuri
1 lot buy GBP/JPY = 1 lot buy GBP/USD + 1 lot buy USD/JPY

This is a typical misbelief by traders who do not think / check it more carefully.

Forget about the $6 / day interest, the fomula itself can be a big killer.
After carefull thought, indeed Samuri is right.

We may open the positions TODAY with 1/1/1 ratio. But after the prices moved up/down several hundred pips away, this ratio is no longer valid.

Example: Open Prices
Dec 1 2006
Long GBPJPY 227.62
Short GBPUSD 1.9658
Short USDJPY 115.80

Jan 2
GBPJPY 233.25 +563pips
GBPUSD 1.9655 +3pips
USDJPY 118.65 -285pips

Feb 1
GBPJPY 237.10 +948pips
GBPUSD 1.9652 +6pips
USDJPY 120.67 -487pips

In the above example, it was fortunate that GBPJPY moved up favoring long GBPJPY, but what if the reverse happend instead?

Though it is true that GBPJPY = GBPUSD * USDJPY, the ratio 1/1/1 is not valid throughout the price spectrum. Interesting, GBPUSD stabilized at 1.9650 and today, its working its best to stay in this price again.

It appears to me that the ratio should be:
1 Long GBPJPY,
1 Short GBPUSD
1.96 Short USDJPY (because GBPUSD is 1.9650 for most part of the period)
By Feb 1,
GBPJPY 237.10 +948pips
GBPUSD 1.9652 +6pips
USDJPY 120.67 -954pips

BUT what if GBPUSD went up to 1.99 last Jan 24 and our ratio is 1/1/1.96
GBPJPY 241.30 +1368pips
GBPUSD 1.9912 -254pips
USDJPY 121.35 -1087pips
Hmm... seems that its still EVEN

So from the above analysis, the ratio for USDJPY should be taken from a very stable GBPUSD average...
  • #8
  • Feb 23, 2007 8:36pm Feb 23, 2007 8:36pm
BK,

You are absolutely right...the ratios are off...and the ratios will fluctuate as the price of the GBP rises and falls.

furthermore, while this may produce a "stable" hedge, the combines negative swaps will more than out-chase the positive swap on the GBPJPY, therefore making this hedge some what of a pointless endeavor.

In case you were wondering, the Positive swap on the GBPJPY is about $25 a day, the GBPUSD short rate is a negative $1.00 and the USDJPY short rate is negative $14...if you were to short 1.9 lots of the USDJPY, your negative swap is $26.60...which is already more than what you'd get from the positive GBPJPY buy.

So, not sure why you'd want to open this type of a hedge unless you were trying to open up a 1/1/1 ratio with the mistake of thinking your hedge would be stable and you were thinking of collecting a modest swap. To me, the 1/1/1 ratio still means you're betting almost completely on the GBP/JPY going higher...so, why not just reduce the position you were going to take in the GBPJPY to something more modest? Or try to work with different correlations.

Quoting BurgerKing
After carefull thought, indeed Samuri is right.

We may open the positions TODAY with 1/1/1 ratio. But after the prices moved up/down several hundred pips away, this ratio is no longer valid.

Example: Open Prices
Dec 1 2006
Long GBPJPY 227.62
Short GBPUSD 1.9658
Short USDJPY 115.80

Jan 2
GBPJPY 233.25 +563pips
GBPUSD 1.9655 +3pips
USDJPY 118.65 -285pips

Feb 1
GBPJPY 237.10 +948pips
GBPUSD 1.9652 +6pips
USDJPY 120.67 -487pips

In the above example, it was fortunate that GBPJPY moved up favoring long GBPJPY, but what if the reverse happend instead?

Though it is true that GBPJPY = GBPUSD * USDJPY, the ratio 1/1/1 is not valid throughout the price spectrum. Interesting, GBPUSD stabilized at 1.9650 and today, its working its best to stay in this price again.

It appears to me that the ratio should be:
1 Long GBPJPY,
1 Short GBPUSD
1.96 Short USDJPY (because GBPUSD is 1.9650 for most part of the period)
By Feb 1,
GBPJPY 237.10 +948pips
GBPUSD 1.9652 +6pips
USDJPY 120.67 -954pips

BUT what if GBPUSD went up to 1.99 last Jan 24 and our ratio is 1/1/1.96
GBPJPY 241.30 +1368pips
GBPUSD 1.9912 -254pips
USDJPY 121.35 -1087pips
Hmm... seems that its still EVEN

So from the above analysis, the ratio for USDJPY should be taken from a very stable GBPUSD average...
  • #9
  • Feb 23, 2007 9:37pm Feb 23, 2007 9:37pm
Quoting mrmikal
BK,

You are absolutely right...the ratios are off...and the ratios will fluctuate as the price of the GBP rises and falls.

furthermore, while this may produce a "stable" hedge, the combines negative swaps will more than out-chase the positive swap on the GBPJPY, therefore making this hedge some what of a pointless endeavor.

In case you were wondering, the Positive swap on the GBPJPY is about $25 a day, the GBPUSD short rate is a negative $1.00 and the USDJPY short rate is negative $14...if you were to short 1.9 lots of the USDJPY, your negative swap is $26.60...which is already more than what you'd get from the positive GBPJPY buy.

So, not sure why you'd want to open this type of a hedge unless you were trying to open up a 1/1/1 ratio with the mistake of thinking your hedge would be stable and you were thinking of collecting a modest swap. To me, the 1/1/1 ratio still means you're betting almost completely on the GBP/JPY going higher...so, why not just reduce the position you were going to take in the GBPJPY to something more modest? Or try to work with different correlations.
I agree that a three way hedge is a looser (for traders). Brokers already thought of it to make life for hedge carry traders miserable. Even with the complexity added, the hedging is still not accurate because GBPUSD cost $10 per pip while GBPJPY and USDJPY cost $8.4 per pip, so last Jan 24, the pip value of the positions is even but the $ value is negative.

Carry Trades are not as easy as it seems. Newbies are the most vulnerable to make such mistakes. Carry Trades works with the principle of "EXTREME PATIENCE". And in that line, I'd rather wait extremely patiently for JPY to fully strengthen - but then again, swap for GBPJPY may be worth just $1 per day when this happens....

Just my luck... the carry trades is already reaching the final destination and I missed the ride.

Many suggests using CHFJPY to hedge GBPJPY... but this is relying on the assumption that GBPCHF is either sideways to bullish... So if that is the case, this hedge is simply long GBPCHF...
  • #10
  • Feb 23, 2007 9:57pm Feb 23, 2007 9:57pm
Quoting fffx
Searching about Hedging strategies, came accross a techinique which seems interesting (i.e. interest positive). The strategy is to to Buy 1 pair GBP/JPY, sell 1 pair USD/JPY and sell 1 pair GBP/USD. In this triangular hedge, the off balance in the units of Yens would be the volatility,which could be encashed on a positive swing, while all the time accruing imterest@ 6 USD per day per standard lot. Could some experienced trader throw light on this and also please let us know what is the maximum pips negative swing they have experienced or should be expected. Thanks
Three way hedge doesnt make any sense (at least to me).....do some research and you will be able to see how it works. Hedging doesnt work for most pairs. Its very tricky........................

Here is a good website that should shed some light on your question http://www.understandmarket.com/index.php?categoryid=55

ENJOY!!!
  • #11
  • Feb 23, 2007 9:57pm Feb 23, 2007 9:57pm
This whole hedge idea is totally ridiculous. This is another trap that broker-dealers design for poor traders. If you notice the negative interest on any pair is always bigger than positive interest. So this is another device to make up the difference just like spread. Frankly, I'd rather they paid me my measly 3-5% interest on my deposit per annum, especially that after all I don't really buy anything, like one buys stocks or mutual funds, but merely put up my money as a collateral for a contract.
What did they say at the beginning of this thread? They can make $6 a day?
That's what 6x365=2190. Ha, if your hedge, whatever it is, is gone wrong, and it will trust me. you can lose that much in couple of bad days if you are unlucky.
No, thank you Sir.
  • #12
  • Feb 24, 2007 1:06am Feb 24, 2007 1:06am
This idea sounded sounded somewhere between crazy and interesting when I first heard about it. So I put this theory to the test in a $5000 demo account beginning 1/22/07. I bought or sold a standard sized lot for each pair.

My equity now stands at $5252., representing a 5% gain. Balance is $5,947 with a $695 negative float.

A couple of key points: 1.) When the COMBINED profit swung positive $200 or more, I would take that profit by closing each trade and re-starting the hedge - since that wouldn't affect the interest collected.

2.) A $5,000 account is too small for 3 standard sized lots. I came fairly close to a margin call near the bottom of GBP/JPY's recent 800 pip decline.

I BELIEVE that a $10,000 account would be comfortably sufficient for this hedge to survive with 3 lots in play.

My broker pays a net of almost $10 per day in interest. So if the fundamentals of this hedge remained in place - on a $10,000 account you would earn about 35% in interest for a year. By removing profits as I did, it could be 50% or better.

Again, I am not endorsing this idea - but merely adding some facts and figures to the conversation. To some people, a 35-50% annual return with little to no excitement would be fantastic.

Additional testing will reveal things that my first 30 days did not - and all bets are off in the event of a nuclear war. But if you have modest goals, little to no screen time available and a $10,000 account (or scale down), I was surprised to see that it actually worked pretty well.

Slap it into a demo account and see for yourself. But of course all brokers have different interest policies. I used a prominent MetaTrader enabled broker.
  • #13
  • Edited 2:11am Feb 24, 2007 1:52am | Edited 2:11am
A/B * B/C * C/A = 1
with lagged market characteristic it would create value above or below 1 a little, so yes this lagged nature could be used for trading (which banks will fill these imbalance values faster than the other I think)

but as for gaining interest let see,
A/B = pay B interest and gain A interest
B/C = pay C interest and gain B interest
C/A = pay A interest and gain C interest
it sounds like zero-sum to me.

here's the math from BK sample

Dec 1 2006
Long GBPJPY 227.62
Short GBPUSD 1.9658
Short USDJPY 115.80
227.62 x 1/1.9658 x 1/115.80 = 0.99991372328650669101392007121431

Jan 2
GBPJPY 233.25 +563pips
GBPUSD 1.9655 +3pips
USDJPY 118.65 -285pips
233.25 x 1/1.9655 x 1/118.65 = 1.0001862083005163983905685334987

Feb 1
GBPJPY 237.10 +948pips
GBPUSD 1.9652 +6pips
USDJPY 120.67 -487pips
237.10 x 1/1.9655 x 1/120.65 = 0.99982843939169881115802128663844

actually even if the pips ratio are off, the value for 1 pips is different for each pair, so its still in balanced state. so all three position above doesn't differ much in term of P/L states.

The result of this, the one who will always profit is still market makers/banks hehe, because they charge 'spread'
  • #14
  • Feb 24, 2007 2:11am Feb 24, 2007 2:11am
Quoting fffx
Searching about Hedging strategies, came accross a techinique which seems interesting (i.e. interest positive). The strategy is to to Buy 1 pair GBP/JPY, sell 1 pair USD/JPY and sell 1 pair GBP/USD. In this triangular hedge, the off balance in the units of Yens would be the volatility,which could be encashed on a positive swing, while all the time accruing imterest@ 6 USD per day per standard lot. Could some experienced trader throw light on this and also please let us know what is the maximum pips negative swing they have experienced or should be expected. Thanks
I spent a couple hours the other day playing with this idea, and finally figured out why it doesn't really work. Using the current prices (GBP/JPY bid is 237.58, USD/JPY ask is 121.03, GBP/USD ask is 1.9634), here's what happens when you buy even amounts of each (let's say 1 standard lot, or 100k):

Buy 100,000 GBP/JPY: We get 100,000 GBP, and we sell 23,758,000 JPY
Sell 100,000 USD/JPY: We sell 100,000 USD, and we get 12,103,000 JPY
Sell 100,000 GBP/USD: We sell 100,000 GBP, and we get 196,340 USD

The GBP positions cancel out, but we have 96,340 USD long and 11,655,000 JPY short. You'll collect $6.65 daily in swap, but since this is effectively just a long USD/JPY trade, you would do better by just buying 1 lot of USD/JPY and earning $13.50 daily in swap (the difference is taken by your broker), and you'd have the same exposure to price fluctuations. To actually hedge the 3 pairs, you need to vary the lot size. In this case, that would mean:

Buy 100,000 GBP/JPY: We get 100,000 GBP, and we sell 23,758,000 JPY
Sell 196,430 USD/JPY: We sell 194,430 USD, and we get 23,773,923 JPY
Sell 100,000 GBP/USD: We sell 100,000 GBP, and we get 196,340 USD

Now we have no position (except for 15,923 JPY, which is pretty negligible and a result of the Bid/Ask spread). We also earn negative swap (-7.14 per day at ibfx rates), which would be zero but the broker takes a commission on swap too. Too bad there's no free lunch....
  • #15
  • Feb 24, 2007 2:19am Feb 24, 2007 2:19am
Yes, agree with the lot size has to be vary. Lot size for each position for 3 way hedge has to be the same amount of used margin. Which will results a different unit size for each position.
  • #16
  • Feb 24, 2007 4:42am Feb 24, 2007 4:42am
You should not look at hedging GBPJPY with GBPUSD +USDJPY on the same broker, but rather on two or even three brokers.

Its very obvious for brokers to know that you are doing swap arbitrage by hedging the long GBPJPY on a swap paying broker with a short GBPJPY on a swap free broker.. Therefore masking the short GBPJPY with a GBPUSD+USDJPY on a second (swap free or a lower swap) broker makes sense.

ie,

LONG GBPJPY on Broker A

Short GBPUSD, Short USDJPY on Broker B
or
divide these shorts on two brokers...
  • #17
  • Feb 24, 2007 9:24am Feb 24, 2007 9:24am
Quoting BurgerKing
You should not look at hedging GBPJPY with GBPUSD +USDJPY on the same broker, but rather on two or even three brokers.

Its very obvious for brokers to know that you are doing swap arbitrage by hedging the long GBPJPY on a swap paying broker with a short GBPJPY on a swap free broker.. Therefore masking the short GBPJPY with a GBPUSD+USDJPY on a second (swap free or a lower swap) broker makes sense.
First of all, if we involved two or three brokers (accounts, that is), not only would this simple concept get complicated - but we'd need substantially more money in order to keep each of the accounts away from margin call territory.

Secondly, while I don't profess to know everything within the realm of brokers, I don't believe my broker is going to care. So what if I've got three medium to long-term trades going on? It happens every day.

It's a big world - and the odds of a broker caring deeply about my three tiny little trades - would seem to be highly overblown.

With the $10,000 (scaleable) account level that I previously mentioned, a single account could accomodate plenty of small daily trades. So even account inactivity would cease to be a concern to brokers.

I expect that this thread will die down again after this flurry of recent posts. But I'll strive to come back in 30 days and present my updated results.

In the meantime, just realize the FACT that this simple, goofy looking 3-way hedge outperformed 95% of Forex traders for the past month. Only time will tell who wins next month.

And I'm not taking sides. I'm just reporting the facts.
  • #18
  • Feb 24, 2007 10:02am Feb 24, 2007 10:02am
Brentmark yoy've been extermely lucky so far. But email us later when you lose all your money in a demo account.
I tested this strategy extensively a little while ago on various pairs. There is a belief as usual that one can find some magical pair combination in some magical ratio to each other so that they will always be eaualized. But that just doesn't happen. Usually they get out of sync every 3-4 weeks and then go deeper and deeper into red. This happens when syddenly one pair refuses to cooperate and acts wild while other just sit there and hardly move. Then they equalize on a new level which incidentally is $600-700 below what it was.
BTW the guy who gave me this sterategy and was swearing by it, now hawks some martingale system on the net called binary equation (another piece of junk).
  • #19
  • Feb 24, 2007 10:14am Feb 24, 2007 10:14am
Quoting irusoh
Brentmark yoy've been extermely lucky so far. But email us later when you lose all your money in a demo account.
No problem. You've got a deal. I'll let you in on the good, the bad and the ugly.
  • #20
  • Feb 24, 2007 1:42pm Feb 24, 2007 1:42pm
Hi all, i've read through this thread not in an interest in actually getting into a three way hedge, but rather understanding how to do the math on a two way hedge. I know this should be a simple thing for those with math-on-their-minds so that is why i ask here. I love math...i'm just having a bit of a go figuring out the formula for the following:

This is just an example, i'm not saying it's right/wrong or even doable...but could the forumla be done for the following? THX !!!!!!

Buy Eur/Usd (include in formula the spread, long-swap, (and correlation??), + anything i missed??)

Buy Usd/Chf (include in formula the spread, long-swap, (and correlation??), + anything i missed??)

what would be the ratio of how many lots to buy of Eur/Usd vs How many to buy of Chf/Usd.

Then would it be recommended to go someplace like the great link posted in a prior post on the daily/weekly correlation and change that figure in the calculation before entereing??

Thanks in advance for any assistance.

Cheers,
Thom
0 traders viewing now