This page has been archived and commenting is disabled.
Decoupling Spread Closed, Or 6/6/6.5
The most recent iteration of the ES-EURJPY decoupling trade spotted on Friday was a little stubborn going into Friday's close. We were almost worried we may lose this guaranteed money maker for a second. Those worries are now gone: this is the 6th time the observed risk-FX docoupling trade has closed in 6 observations, and in 6 (and a half) trading sessions. With cumulative P&L over the past week starting to become material, on a standalone, unlevered and/or annualized basis, we are stunned that this trade continues to manifest itself with such regularity.
»
- 4200 reads
- Printer-friendly version
- Send to friend
- advertisements -
Spread, bitchez
Hmm. Next weeks shaping up to be a strange one for currency.
Lol this blog is hilarious. Are you looking at the interpolated close to open prices for the spoos and the 200 tick wide weekend market in the EURJPY to reach your conclusion?
*plonk*
imposter!
No, just looking at what passes for an efficient, non-cherrypicked market
i'm confused again, Tyler, perhaps you can help. your previous post
http://www.zerohedge.com/article/market-decoupling-world-cup-edition-or-...
had the starting levels ESU0: 1077.75 EURJPY: 110.59, with the chart clearly showing the ESU was outperforming, and would thus be shorted vs the FX portion.
this post has the closing levels: ESU0: 1091.25 EURJPY 111.81
so, the PnL is -125bps on the ESU portion, and +110bps on the EURJPY trade. for a net of -15bps.
i guess you make up for it in volume?
Reindex to 100% baseline. roughly 10x leverage in FX side
i'm using your starting numbers and your ending numbers - reindexing has nothing to do with it.
if you lever up the FX side 10x, you're not making a convergence bet, you're just making an FX bet... no?
reindexing to a 100% baseline from a converged inception point anticipating a convergence, is a convergence bet, and when convergence happens, you profit. The trade can be structured in a hundred different ways.
lol. and yet, i'm using the numbers from your starting post, and the number from your ending post. and there is no profit. of course, that is, until after the fact when you say "oh - the FX side just has to be much bigger," which of course means that the total PnL depends on the performance of the FX trade, and not on the actual convergence.
To make a profit you'll need to master at least 4 advanced math concepts:
i see. i'm waiting for anyone to point out which one of those 4 operations i got wrong...
You need to balance the legs. Like TD said before there multiple ways to structure the trade.
[you either junk or comment; a junk is not worth commenting]
good point about the junk/comment. i had a change of heart, sorry.
look - in a convergence trade, where you're betting on the relative performance of two assets like this, ESPECIALLY when you try to construct graphs that show how the performance converges from a divergence off of an initially indexed value of 100, the notionals are equal. That's the entire point of using the fancy graphs that show divergence and then convergence based on an initially indexed value.
surprisingly, i'm the only one of TD's millions of readers that noticed (or cared) that he re-indexed from DIFFERENT starting points in his initial post and his closing post - i'm sure it wasn't an accident - he needed to do that to make the convergence look good, because the numbers simply didn't converge.
in reality, you can look at the NUMBERS, which are in each post, and which I did the very simple math on above, and see that there in fact was no convergence.
on the other hand, the EURJPY did rally, so if you just made a big bet on EURJPY, which is what tyler suggested after the fact, and has nothing to do with convergence at all - you made money.
I don't know were you got this from but it is a wrong assumption and the source of all your confusion.
That should be a warning sign to yourself. When you are the only one in a specialized site of any matter, you are probably wrong.
I have to move on but I hope I helped somehow.
thanks for the effort, Nonclaim, but no. If we want to play Tyler's "oh, the FX side is 10x larger" game, then this post:
http://www.zerohedge.com/article/and-scene-4-out-4-4-days
results in massive losses, as the convergence was to the downside. Can't have it both ways.
no i noticed it to...when i said why wouldnt you just go short eur/yen and profit from that trade.
must say i agree with kid dyno-mite.
but i got a reponse trying to talk over my head to make me look stupid. at least i keep my variables consistent.
I guess this means the market will be going higher. The pump machine was fixed last week. Time to go long. Is this really surprising? The Fed has the most powerfulsoftware that is designed to stomp out all us shorts. Why do you toture yourselves . Go with the flow and stop analysing.
Can someone (simply) explain how to make money on this trade?
Please!
short 1 es contract
buy a dollar equivalent amount of eurjpy
close both sides when the lines cross again
to be even more clear:
1 es contract is worth ~ $54,000.
45,000 eurjpy is worth Y 5,000,000 ~ $55,000
entering at 1pm EST Friday, you would have shorted ES at 1080, and gone long the eurjpy at 110.57. You'd have exited this evening at 9:20 EST, at 1090 and 111.86
except that if you use the actual data points from the actual posts that Tyler is using to show this supposed moneymaker, the numbers don't work.
DOW chart warns of a rally :
http://stockmarket618.wordpress.com
http://www.zerohedge.com/forum/latest-market-outlook-1
Kid
Your maths should show at least a 10pip profit (not great on this trade but most have averaged at least 20-30pips).
had the starting levels ESU0: 1077.75 EURJPY: 110.59, with the chart clearly showing the ESU was outperforming, and would thus be shorted vs the FX portion.
this post has the closing levels: ESU0: 1091.25 EURJPY 111.81
1091.25-1077.75 = 13.5pts - commissions
111.81 - 110.59 = 122 pips - commissions
135-122 = +13pips and take off a few for commissions.
TBH, the rule is that when you find an arb, you keep quiet about it or it will fail so whilst I'm happy (like others) that this has been shown, I think we should shut up about it :)
lol...and turn into squid jr. no wonder our world is fucked.
BG - huh? you can't convert ESU points to EURJPY pips, but anyway you have the sides wrong. in your example it would be 122-135 = -13
look - if you shorted ESU you lost 125 bps on whatever notional you did it on, and if you bot EURJPY you made 110bps on whatever notional you did it on.
agreed on that much at least?
then, tyler, commenting above, now says that you have to do 10x as much on the FX side... that's not a convergence trade, it's just an FX bet...
i think the problem is that if you want to show this divergence in chart form, you need to graph both assets on the same graph with the X-axis as time, the Y-axis as % change, and the starting point of the trade indexed to 0 on the Y axis. Normally, Tylers chart's should show something equivalent, although for some reason his two Y-axis are not scaled equally!!!
in other words, it's not that the ESU trades at 10x the value of the EURJPY that's the problem -it's that the scale is not consistent. A 1% change in EURJPY does not result in the same line as a 1% change in the ESU on his chart - and that is precisely why the data is funked up.
you would be better off just going long POPN when it trades back to .0011.
yeah thats right. you heard it hear first.
got it now - in the wake of my comment below about how the scale (not the values, but the scale) of the two y-axes are different... you need to lever the FX position by the difference in the two scales. it's roughly 1.5 times. the mystery to me now is why TD said 10x.