Safer way to trade the carry trade?

[PeterPG]

hey guys,

I have been interested in carry trades for a long time now and been looking at ways to trade them with less risk, as traded them naked is too risky for my risk appetite.

I looked at the charts for NZD/JPY and TRY/JPY. Historically they have been very closely correlated, even in the recent major unwinding. I am currently trying a long TRY/JPY position and hedging with a short NZD/JPY. TRY is the turkish lira and the base rate is 17.5%. The strategy would give quite a nice margin for error if held medium to long term.I am using Saxobank for the trade as they have both try/jpy and nzd/jpy. IT could be used with as much leverage as you felt was best suited to you. I am very conservative by nature. Any thoughts ?

 

[LION63]

hey guys,

I have been interested in carry trades for a long time now and been looking at ways to trade them with less risk, as traded them naked is too risky for my risk appetite.

This has become a very loose term and most do not understand what it really means. A CARRY TRADE IS SIMPLY BORROWING MONEY AT X% INTEREST AND INVESTING THE MONEY IN AN ASSET THAT PAYS X+Y. By definition this means that the trade is absolutely risk free and all the trader needs to ensure is that interest rates do not rise and the asset yield does not fall.

The exotic play of borrowing in one currency to invest in assets of another currency are just an add on which markedly increases the risk to the trader.

 

[grantx]

Peter,

Sounds like one for Dashing Blade but in the meantime, here’s my modest contribution.

Your net position is therefore long TRY (high yield), short NZD (high yield), zero JPY (no yield - long JPY plus short JPY). Also, you pay interest on the short high yield and receive interest on the long high yield. Long/short high yield = no yield. I think this may have immunised (hedged) the intention. I stand to be corrected.

Grant.

 

[jacinto]

hey guys,

I have been interested in carry trades for a long time now and been looking at ways to trade them with less risk, as traded them naked is too risky for my risk appetite.

I looked at the charts for NZD/JPY and TRY/JPY. Historically they have been very closely correlated, even in the recent major unwinding. I am currently trying a long TRY/JPY position and hedging with a short NZD/JPY. TRY is the turkish lira and the base rate is 17.5%. The strategy would give quite a nice margin for error if held medium to long term.I am using Saxobank for the trade as they have both try/jpy and nzd/jpy. IT could be used with as much leverage as you felt was best suited to you. I am very conservative by nature. Any thoughts ?

long TRY/JPY = +TRY and a - JPY
Short NZD/JPY = - NZD and a + JPY

in other words

Long, TRY/NZD

 

[timm]

hey guys,

I have been interested in carry trades for a long time now and been looking at ways to trade them with less risk, as traded them naked is too risky for my risk appetite.

I looked at the charts for NZD/JPY and TRY/JPY. Historically they have been very closely correlated, even in the recent major unwinding. I am currently trying a long TRY/JPY position and hedging with a short NZD/JPY. TRY is the turkish lira and the base rate is 17.5%. The strategy would give quite a nice margin for error if held medium to long term.I am using Saxobank for the trade as they have both try/jpy and nzd/jpy. IT could be used with as much leverage as you felt was best suited to you. I am very conservative by nature. Any thoughts ?

I don't have historical data for TRY to hand, but I had a quick look on a chart. The Turkish Lira seems to be a less popular investment for the 'japanese housewife', so is moved less by the JPY carry traders. This means that TRY/JPY has been less volatile than NZD/JPY. So a long position on TRY/JPY and a smaller short position on NZD/JPY maybe could have cancelled out most of the recent volatility. AUD could also be used in this strategy.

Of course, currency correlations are always changing, so you can never really hedge out the risk this way. You must have a stop to limit risk and a complex position like this is more difficult to put a stop-loss on.

Exotic currencies like TRY are also not commonly associated with conservative low-risk traders. Be prepared for big moves due to political events. It all sounds very scary to me.

 

[timm]

Primarily AUD, NZD and to some extent EUR are the main currencies to see Uridashi related activity.

Thanks GJ. It certainly helps to know the fundamental reason for these correlations if you are going to bank on them.

Is there a reason for these particular currencies to be favoured? And is Uridashi activity in other currencies increasing?

 

[amit_rwd]

while trading try/jpy and nzd/jpy you have to watch the liquidity and volatility of both high yields. Try is less liquid then nzd and the trading hours of try are limited so in the event crisis stuck and major unwinnding takes place in asian session you will be stuck with try/jpy until european open. Secondly the volatility of try is higher than nzd and as per your analaysis you found very close corelation between try/jpy and nzd/jpy but thats because 1> emprically during crisis all assets move up or down correlation is almost 1 ii> i bet in % terms try/jpy suffered more downfall. so you are correct in terms of getting a partical hedge but remember when stocks suffer losses try will be affected the most due to carry trade( +ve) corelation with s&p and dow .
other possible carry trades i can recommend which are not volatile try/dkk = try/usd +usd/dkk (danish krone) or try/chf = try/usd+ usd/chf
you can use sgd(singapore dollar) and hkd (hongkong dollar ) for funding instead of jpy but check the trading times...
if you are creating carry trade portfolio then make sure you exposure to usd in terms of short and long doesnt have wide gap
say for examplein a 100k portfolio you are 70k short usd in carry trade then make sure you are 30k long in usd so you dont have to suffer one sided dollar strenght moves..

hope its helps and if doesnt them pm me.....

 

[ortizfabio]

Is there anyway to do this everyday before the close of the day (5:00 PM for IB) and then exit the trade as soon as posible (When the asian/european market open) with a limit order? This way you can make the profit of the interest differential and cut down the exposure to currency movement? Also it could be a good way to put your money to work every night until the US market opens next day. The only risk I see is if any of the currencies goes far against you from what you entered you might have take the loss of hold it until it goes your way. Also it could be better to use all asian currencies for me being in the US. This way the gap between entry/exit will be narrowed but again one might not be able to enter because there might not be quotes for all asian currencies at 5:00 PM. I had one instance where I was in a trade with NZD/JPY and I wanted to exit at 1:00 PM and I could not because IB was not quoting this currency.


Thanks,

 

[amit_rwd]

hi ortizfabio you cant do this because of emerging market currency spreads are high and if you follow the strategy you describe u will make terrible losses.. and possibly double whammy's if position is against you... the best way is to make your own small portfolio in oanda ... but right now market conditions dont favour carry trades atall.. so wait and watch when economies are in healthy state then go for it..

 

[Phoenix669]

theres no risk free way to hedge the carry trade on retail unless you sign up for a non interest paying/receiving brokerage account which often has steep rules as to what you can do with the account.

This was my USD/ISK engineering hedge, hasn't been proven, but could be plausible but only at institutional level.

Take a foreign currency loan from a lower interest rate country (i.e Japan @ .5%, said japan just to maximize the trade). In a Foreign currency loan, you burrow say usd from the bank of japan's reserve. Then you would take a short cash USD/ISK position, and buy Govt bonds. Now to hedge this as you only need to pay back usd and not jpy, you can buy a Forward contract on USDISK. The profit difference is:

FV=PVe(R3(R2-R1)(T)
PV= amount burrowed
R3=ISK interest rate
R2=USD interest rate
R1=JPY interest rate
T=time

now as I am bullish on USD interest rates, the USDISK forward will also make a little more money as interest rates go up because the gap between R2 and R3 is getting smaller, and the gap between R3 is larger.

Might have made a few mistakes, open to opinions on this.

 

[ortizfabio]

Not so sofisticated

Phoenix669,

I am going to write what I got from you post.

Lets say I borrow $100,000 USD from the bank of Japan then I open an account in IB and do the following. Short the USD/ISK, would that be a position of $100,000 USD? if so then I end up with 12,034,086.44 ISK as balance. (notice I have not use any leverage).
Now I am not sure how to buy the US Bonds and what is the symbol for the bonds. If I used all my money to short the USD/ISK how I am going to buy the us bonds?

So let's plug the values in your formula for a period of a month (20 days) and see what happens.

FV=PVe(R3(R2-R1)(T)

FV = (100000 * (0.115 * (0.02 - 0.005))) 20
FV = 3450 USD

so for a year it would be 41400

Now the million dollar question, How do I go about getting $100,000 USD from the bank of japan. Oh I remmember now that you said "but could be plausible but only at institutional level."

How I am doing so far?

Thanks,

Fabio :smart:

 

[Phoenix669]

O.k so let me explain it a little better.

You are a hedge fund manager. You want to create a rather substantial risk free rate for your firm, and are looking to invest 1 000 000 USD. Seeing how your institutional, you have more tools to your palate.

You begin to search for a place to leverage your cash flow and find out you can take foreign currency loans from the central bank of a country at that countries current interest rate. You take your 1 000 000 USD to that country (for this example Japan) and leverage it to 10 000 000 USD @ 0.5% interest.

You now have 10 000 000 USD, not JPY.

You change your 10 000 000 USD into ISK. Which (using xe.com) is 813,008,759.62 ISK.

You then proceed to buy ISK bonds (Bonds) at 12.43% for say 1 year (for sake of argument rounding to 15%).

Now because you are burrowing USD from BOJ at 0.5% you only need to pay back USD.

You Decide to hedge your position because you are hawkish on USD interest rates, so you buy a long forward contract on USD/ISK. Basically locking it in at the same rate you exchanged at (keeping in mind that you only need to pay back USD not JPY). This creates arbitrage position on the interest rates.

Now to calculate terms on the future value of the transaction (not including cost of forward) you have this:

(used wrong equation in first post) actual equation is: FV=PV*(1+R3-(r2-r1))^T

PV= 10 000 000 (813,008,759.62 ISK)
R3= .15
R2= .02
R1= .005
T= 1

so you have

FV=10 000 000*(1+0.15-0.02-0.005)
FV=10 000 000*(1.125)
FV=11 250 000

Now you need to repay 10 000 000 so your profit is:

P=11 250 000(profit from leverage) - 10 000 000(amount burrowed)
P=1 250 000usd

hopefully that made more sense.

 

[ortizfabio]

Phoenix669:

Very interesting carry trade unfortunatelly it looks like is not doable for a mere mortals. Only hedge fund manager Gods could get a loan from the central bank of Japan. The logistics appear complicated like where to buy ISK bonds. The ROI is really awesome 125% yearly. Not sure where the risk is. You said to buy the forward contract so that means some money needs to be allocated there therefore the ROI would be reduced by whatever the cost of the forwards is.

By the way nice blog I wish I had started at your age.

Thanks,

Fabio

 

[Phoenix669]

thanks...I had a link posted to the official ISK bonds website here Bonds

 

[£10kLoser]

Why make it so complicated?

Just buy the rolling daily GBP/JYP (for example) with Capital Spreads and sell the GBP/JYP future with them or preferably another spread bet co. The former pays interest, the latter charges no interest. In less than a week, you've covered the spread and thereafter get the interest paid daily with zero risk since one trade exactly counters the other.

This really is free money and you'll get about 30% on the capital required to cover the losing side (needed for both trades since we don't know which will lose).

 

[Phoenix669]

futures are discounted due to the time value of money; the interest over time...unless you have a spot-futures arb or future-spot convergence (only in expiry month) then you are neutral interest rates.

Not to mention interest rate differentials between the brokers...rollover for retail isn't the exact rollover, so you have to take partials to match it up exactly.

 

[£10kLoser]

Say That Again?

futures are discounted due to the time value of money; the interest over time...unless you have a spot-futures arb or future-spot convergence (only in expiry month) then you are neutral interest rates.

Not to mention interest rate differentials between the brokers...rollover for retail isn't the exact rollover, so you have to take partials to match it up exactly.

The only people who could understand that gobbledegook are those who knew it already...

 

[ortizfabio]

£10kLoser:

You got my attention, finally a simple way to make some extra money. Would care expanding a bit your explanation? For example what symbols specifically to trade? "rolling daily GBP/JYP with Capital Spreads " Are you talking about buying the FX spot pair GBP/JPY? otherwise what symbol is that? How long to hold those symbols. My idea was to find a way to make some money at the end of the trading day but I would need to have my margin back for the next intraday session (930-1600 EST). Is that possible with your setup?

Thanks,

Fabio

 

[£10kLoser]

Carry Trade

Hi


Yes, FX spot pair. the theory behind the carry trade is that you buy currency with the higher interest rate. so, pick a high interest rate paired against a low interest rate currency and you will be paid the difference daily.

However, the price move on the pair will obviously be much greater, so you need to sell the pair somewhere where there is no interest rate charges, ie. the quarterly future.

This has to be a long-term startegy (weeks) and your capital will be tied up since you need to cover the losses on the losing trade (which will, however, be balanced by the winning trade).

The future and the spot price have a divergance depending on how far away from the expiry of the quarterly future. For example, spot GBP/JYP is currently 205.36 and the September future is 204.67 - this converges as the expiry gets closer. The Dec future is 202 something - indicating that the price is expected to fall.

This discrepancy stops the system being 100% guaranteed to make a profit. For 100% guaranteed, you need to find a broker that doesn't charge an overnight swap and sell the pair with them. However, the risk there is that such brokers tend to be less reliable!


So, as Pheonix669 pointed out, there are problems with this system if you implement it by hedging with the future.