Please allow Macro Man to apologize for the rather rude and witless title to today’s post. He generally prefers to conduct himself in a (relatively) polite and professional manner. However, the subject of today’s post, Dr. Nouriel Roubini, has developed the charming habit of calling those who disagree with his analysis “fools”. What better way, then, to refute the good doctor than by starting off one’s argument with an epithet taken straight from the schoolyard?
In any event, Dr. Roubini and his merry band of globe-trotting pundits at RGE have decided that the yen carry trade is the latest issue that could undermine the global financial system and potentially send the world careening into recession. Having missed the boat with housing and the economy (yes, the housing market has come off sharply, as accurately forecast by RGE; no, this has not produced a recession, as forecast by RGE), they have decided to invoke the spirit of 1998.
In fairness, Roubini is not the only one. Macro Man has seen and heard an increasing drumbeat (mostly from non-currency specialists) over the past several months about the size and danger posed by the yen carry trade, and how it risks a catastrophic unwind a la 1998. Indeed, the recent World Economic Forum at Davos was full of angst and hand-wringing from assorted worthies about the endgame for the yen.
Michael Lewis did a fair job of dismissing the general doom and gloom of Dr. Roubini and the Davos Doom-Mongers (sounds like a 60’s novelty band, doesn’t it?). The good doctor rewarded Lewis by calling him a “reckless fool”, an especial token of esteem from all and sundry at RGE. Macro Man now proposes to discuss the specific argument of the yen carry trade; the conclusion, as if you couldn’t guess, is that the carry trade ain’t what it used to be.
The “1998 all over again” arguments appear to hinge on two factors. One is that the yen is at “20 year lows” in real terms, which by definition appears to mean that the world is funding out of JPY. The second looks at positioning on the IMM in Chicago, which admittedly is at a record extreme, and presumes that this reflects positioning for the currency market as a whole.
Valuation
Japan’s Vice-Minister for International Affairs, Hiroshi Watanabe, has disputed that the yen is as cheap as standard models suggest. Although the yen is indeed very cheap on a real effective exchange rate (REER) basis, a substantial degree of this cheapness comes from Japan’s deflation of the past decade. Standard calculations suggest that the yen should strengthen during periods of deflation to maintain a constant REER. Watanabe suggests that this is nonsense, and that the appropriate policy response during deflation should be for the yen to weaken, thus helping to stem deflation. While his comments may not hold academic water, they do resonate from a common sense perspective.
Regardless, if the yen were as cheap as REER analysis suggests, we should expect to see that reflected in Japan’s trade account. With Japanese domestic demand tepid and growth strong in most of its trading partners, surely yen weakness has prompted record trade surpluses in Japan? Uh-uh. In fact, Japan’s trade surplus is currently smaller than the average of the past 20 years. This surely suggests that the underlying premise, that the yen’s REER accurately reflects its “true” valuation, is mis-placed.
Positioning
The obvious indicator here is the record short position in yen futures contracts on Chicago’s IMM exchange. It is certainly the case that this indicator is at an historical extreme, and has been for some time. And if currencies were only traded on the IMM, it clearly would be cause for concern vis-Ã -vis the yen carry trade.
Bloomberg
However, the IMM represents only a very small sub-set of foreign exchange traders. The people who use the IMM are typically momentum-driven commodity trading advisors, an important yet nonetheless non-hegemonic participant in the currency market. Very few hedge funds or real money investors use the IMM; clearly, their positions, which are generally more reflective of “fundamentals”, can differ dramatically from those of purely price-based CTAs.
In fact, that appears to be the case. While hedge fund positioning is notoriously difficult to get ahold of, there are a number of surveys of real money equity and bond managers that can provide a reading on positioning from more ‘fundamentally’ based managers. And given that many real money shops now operate hedge funds, these surveys perhaps provide a reasonable glimpse of what hedge fund positioning is like as well.
One such survey is the Deutsche Bank/Russell Mellon survey. And the message there couldn’t be clearer: 2007 is nothing like 1998. In 1998, the median fixed income manager was very significantly short of yen; indeed, even the 25th percentile manager had a small yen short on. Fast forward to today. While there is a cohort of the market that clearly has the yen carry trade on (as indicated in the 75th percentile line in the chart below), the median manager’s yen position is only very slightly short, having been long or flat for most of the past several years. The 25th percentile manager, meanwhile, retains a yen long.
Deutsche Bank/Russell Mellon
It seems clear, therefore, that the yen carry trade is neither as large nor as broad as it was in 1998. Anyone who cites the IMM data as “proof” of FX market positioning is almost certainly an academic or an amateur.
Another way to gauge positioning is to look at Japanese banks’ yen lending to foreigners. Currency traders must ultimately borrow yen when they sell it short, and this borrowing generally filters bank into Japanese banks, with whom other banks have their yen nostros. We can therefore get a read on yen borrowing by looking at Japanese banks’ yen lending.
As the chart above illustrates, there has been a substantial increase in Japanese bank lending to foreigners...which has almost taken it back to where it was sixteen years ago. This data suggests that the apex of the yen carry trade was not in fact 1998, but 2002, when USD/JPY traded up 135 before gently declining over the next several years (with no massive repercussions for the global financial system, funnily enough.) While the recent sharp increase in lending does suggest a carry-trade buildup of sorts, the fact that lending has already rolled over a touch is far from indicative of risk appetite gone wild.
Finally, among the most enthusiastic sellers of yen have been the Japanese themselves. While it is true that retail currency trading volumes have exploded, and could perhaps pose a small risk, the overall size of these positions is very small compared with the currency market as a whole. More persistent is the retail appetite for foreign currency investment vehicles. These include foreign currency uridashi bond funds, but also to an increasing degree, foreign equity funds. Japanese households retain more than 50% of their financial assets in yen deposits that pay little more than zero. With the normalization of the economy and concomitant increase in risk appetite, it is small wonder that Japanese retail has lowered its home bias to a measurable degree.
Of course, when Mr. and Mrs. Jones buy foreign equities (as they did in record volumes last year), it is a sign of loss of faith in the American economy and financial system-and therefore a prudent move. When Mr. and Mrs. Kobayashi do the same thing, it is a sign of reckless risk appetite and the carry trade gone mad. Why? Well, because the good Dr. Roubini says so, of course!
In any event, Dr. Roubini and his merry band of globe-trotting pundits at RGE have decided that the yen carry trade is the latest issue that could undermine the global financial system and potentially send the world careening into recession. Having missed the boat with housing and the economy (yes, the housing market has come off sharply, as accurately forecast by RGE; no, this has not produced a recession, as forecast by RGE), they have decided to invoke the spirit of 1998.
In fairness, Roubini is not the only one. Macro Man has seen and heard an increasing drumbeat (mostly from non-currency specialists) over the past several months about the size and danger posed by the yen carry trade, and how it risks a catastrophic unwind a la 1998. Indeed, the recent World Economic Forum at Davos was full of angst and hand-wringing from assorted worthies about the endgame for the yen.
Michael Lewis did a fair job of dismissing the general doom and gloom of Dr. Roubini and the Davos Doom-Mongers (sounds like a 60’s novelty band, doesn’t it?). The good doctor rewarded Lewis by calling him a “reckless fool”, an especial token of esteem from all and sundry at RGE. Macro Man now proposes to discuss the specific argument of the yen carry trade; the conclusion, as if you couldn’t guess, is that the carry trade ain’t what it used to be.
The “1998 all over again” arguments appear to hinge on two factors. One is that the yen is at “20 year lows” in real terms, which by definition appears to mean that the world is funding out of JPY. The second looks at positioning on the IMM in Chicago, which admittedly is at a record extreme, and presumes that this reflects positioning for the currency market as a whole.
Valuation
Japan’s Vice-Minister for International Affairs, Hiroshi Watanabe, has disputed that the yen is as cheap as standard models suggest. Although the yen is indeed very cheap on a real effective exchange rate (REER) basis, a substantial degree of this cheapness comes from Japan’s deflation of the past decade. Standard calculations suggest that the yen should strengthen during periods of deflation to maintain a constant REER. Watanabe suggests that this is nonsense, and that the appropriate policy response during deflation should be for the yen to weaken, thus helping to stem deflation. While his comments may not hold academic water, they do resonate from a common sense perspective.
Regardless, if the yen were as cheap as REER analysis suggests, we should expect to see that reflected in Japan’s trade account. With Japanese domestic demand tepid and growth strong in most of its trading partners, surely yen weakness has prompted record trade surpluses in Japan? Uh-uh. In fact, Japan’s trade surplus is currently smaller than the average of the past 20 years. This surely suggests that the underlying premise, that the yen’s REER accurately reflects its “true” valuation, is mis-placed.
PositioningThe obvious indicator here is the record short position in yen futures contracts on Chicago’s IMM exchange. It is certainly the case that this indicator is at an historical extreme, and has been for some time. And if currencies were only traded on the IMM, it clearly would be cause for concern vis-Ã -vis the yen carry trade.
BloombergHowever, the IMM represents only a very small sub-set of foreign exchange traders. The people who use the IMM are typically momentum-driven commodity trading advisors, an important yet nonetheless non-hegemonic participant in the currency market. Very few hedge funds or real money investors use the IMM; clearly, their positions, which are generally more reflective of “fundamentals”, can differ dramatically from those of purely price-based CTAs.
In fact, that appears to be the case. While hedge fund positioning is notoriously difficult to get ahold of, there are a number of surveys of real money equity and bond managers that can provide a reading on positioning from more ‘fundamentally’ based managers. And given that many real money shops now operate hedge funds, these surveys perhaps provide a reasonable glimpse of what hedge fund positioning is like as well.
One such survey is the Deutsche Bank/Russell Mellon survey. And the message there couldn’t be clearer: 2007 is nothing like 1998. In 1998, the median fixed income manager was very significantly short of yen; indeed, even the 25th percentile manager had a small yen short on. Fast forward to today. While there is a cohort of the market that clearly has the yen carry trade on (as indicated in the 75th percentile line in the chart below), the median manager’s yen position is only very slightly short, having been long or flat for most of the past several years. The 25th percentile manager, meanwhile, retains a yen long.
Deutsche Bank/Russell MellonIt seems clear, therefore, that the yen carry trade is neither as large nor as broad as it was in 1998. Anyone who cites the IMM data as “proof” of FX market positioning is almost certainly an academic or an amateur.
Another way to gauge positioning is to look at Japanese banks’ yen lending to foreigners. Currency traders must ultimately borrow yen when they sell it short, and this borrowing generally filters bank into Japanese banks, with whom other banks have their yen nostros. We can therefore get a read on yen borrowing by looking at Japanese banks’ yen lending.
As the chart above illustrates, there has been a substantial increase in Japanese bank lending to foreigners...which has almost taken it back to where it was sixteen years ago. This data suggests that the apex of the yen carry trade was not in fact 1998, but 2002, when USD/JPY traded up 135 before gently declining over the next several years (with no massive repercussions for the global financial system, funnily enough.) While the recent sharp increase in lending does suggest a carry-trade buildup of sorts, the fact that lending has already rolled over a touch is far from indicative of risk appetite gone wild.Finally, among the most enthusiastic sellers of yen have been the Japanese themselves. While it is true that retail currency trading volumes have exploded, and could perhaps pose a small risk, the overall size of these positions is very small compared with the currency market as a whole. More persistent is the retail appetite for foreign currency investment vehicles. These include foreign currency uridashi bond funds, but also to an increasing degree, foreign equity funds. Japanese households retain more than 50% of their financial assets in yen deposits that pay little more than zero. With the normalization of the economy and concomitant increase in risk appetite, it is small wonder that Japanese retail has lowered its home bias to a measurable degree.
Of course, when Mr. and Mrs. Jones buy foreign equities (as they did in record volumes last year), it is a sign of loss of faith in the American economy and financial system-and therefore a prudent move. When Mr. and Mrs. Kobayashi do the same thing, it is a sign of reckless risk appetite and the carry trade gone mad. Why? Well, because the good Dr. Roubini says so, of course!
Three other issues
Of course, some of the doom-mongers can provide other ‘evidence’ that their vindication is right around the corner. Chief among these factors is the impending rate rise from the BOJ, which in fact has been impending for about eight months now. While it is true that a BOJ rate hike may provide a temporary bout of support for the yen, it is unlikely to produce a large scale wash-out of short yen positions. Even if the BOJ were to hike 0.25% or 1.25%, the yen would be far and away the lowest yielding currency in the world (outside of some NDF currencies that occasionally price in negative rates!) and yield a substantial discount to other major currencies. Just as the CHF has failed to benefit from SNB tightening (since the SNB started raising interest rates from 0.25% to 2.00%, EUR/CHF has rallied from 1.50 to 1.62), Macro Man believes that any BOJ rate below, say, 1.5% will not be sufficient to dissuade carry traders.
Amazingly, USD/JPY is more or less right where it ‘should’ be on the basis of pre-quantitative easing relative monetary base growth!
Moreover, assertions that the fundamental backdrop to the short USD/JPY are ‘just like 1998’ or even 2002 are well wide of the mark. The last two episodes of yen weakness were characterized by a ‘sell Japan’ mentality, which evinced itself not only in currency shorts but also in equity shorts.
In 1998, for example, Japanese banks were in such bad shape that their yen borrowing rates were substantially higher than non-Japanese banks. This ‘Japan premium’, represented by the spread between 3 month TIBOR and 3 month LIBOR, no longer exists.
Bloomberg
Finally, it seems likely that international equity investors are positioned differently from previous episodes of yen weakness. The last two episodes of yen weakness (1998 and 2001-02) were accompanied by substantial weakness in Japanese equities. Although firm positioning data for international equity index managers is difficult to come by, Macro Man has a strong anecdotal recollection that both real money and hedge fund managers were short/underweight Japanese equities at the time.
In other words, in addition to yen shorts from the carry trade, there was another layer of yen shorts from underinvested/short equity managers. Today, this is not the case. The Nikkei has been one of the darlings of Macroland for several years now, albeit with disappointing performance in 2006. As such, any signs of improvement in Japan will not necessarily be met with yen buying from short covering equity managers. As such, a key factor that made prior yen squeezes so vicious is notably absent today.
Ultimately, there is one category of ‘investor’ who has the yen carry trade on in enormous size: Macro man’s favourite financial market participant, the FX reserve managers. According to the latest IMF data, which is admittedly un-comprehensive, the yen makes up just 3% of the reserve baskets of those central banks who deign to disclose their reserve weightings. Sterling, on the other hand, has a 4% weighting (and nearly 6% amongst emerging economy central banks.) If that’s not the carry trade, what is?
Ultimately, the yen should and probably will remain weak on a broad basis until one of two things happens.
If Japan careens back into deep recession, Mrs. Kobayashi will likely curtail her purchases of foreign securities, eliminating a substantial headwind for the yen. At that point, risks of a short-covering rally would increase.
On the other hand, Voldemort and his cronies could finally alter their behaviour and start buying yen like they do euros. If and when that happens (and it might not be that far off), the yen will rise very substantially, and it will be a great trade to be long yen against everything.
For the time being, however, neither of these factors is in play. And while the good Dr. Roubini jets around the world (presumably on someone else’s dime) wringing his hands with worry, those of us who have real P/Ls will remain more concerned with finding out the real story than the interesting story. And if that makes us fools, well, then so be it. I’d rather be a fool than a big fat idiot any day.
Of course, some of the doom-mongers can provide other ‘evidence’ that their vindication is right around the corner. Chief among these factors is the impending rate rise from the BOJ, which in fact has been impending for about eight months now. While it is true that a BOJ rate hike may provide a temporary bout of support for the yen, it is unlikely to produce a large scale wash-out of short yen positions. Even if the BOJ were to hike 0.25% or 1.25%, the yen would be far and away the lowest yielding currency in the world (outside of some NDF currencies that occasionally price in negative rates!) and yield a substantial discount to other major currencies. Just as the CHF has failed to benefit from SNB tightening (since the SNB started raising interest rates from 0.25% to 2.00%, EUR/CHF has rallied from 1.50 to 1.62), Macro Man believes that any BOJ rate below, say, 1.5% will not be sufficient to dissuade carry traders.
Amazingly, USD/JPY is more or less right where it ‘should’ be on the basis of pre-quantitative easing relative monetary base growth!
Moreover, assertions that the fundamental backdrop to the short USD/JPY are ‘just like 1998’ or even 2002 are well wide of the mark. The last two episodes of yen weakness were characterized by a ‘sell Japan’ mentality, which evinced itself not only in currency shorts but also in equity shorts.
In 1998, for example, Japanese banks were in such bad shape that their yen borrowing rates were substantially higher than non-Japanese banks. This ‘Japan premium’, represented by the spread between 3 month TIBOR and 3 month LIBOR, no longer exists.
Bloomberg
Finally, it seems likely that international equity investors are positioned differently from previous episodes of yen weakness. The last two episodes of yen weakness (1998 and 2001-02) were accompanied by substantial weakness in Japanese equities. Although firm positioning data for international equity index managers is difficult to come by, Macro Man has a strong anecdotal recollection that both real money and hedge fund managers were short/underweight Japanese equities at the time.
In other words, in addition to yen shorts from the carry trade, there was another layer of yen shorts from underinvested/short equity managers. Today, this is not the case. The Nikkei has been one of the darlings of Macroland for several years now, albeit with disappointing performance in 2006. As such, any signs of improvement in Japan will not necessarily be met with yen buying from short covering equity managers. As such, a key factor that made prior yen squeezes so vicious is notably absent today.Ultimately, there is one category of ‘investor’ who has the yen carry trade on in enormous size: Macro man’s favourite financial market participant, the FX reserve managers. According to the latest IMF data, which is admittedly un-comprehensive, the yen makes up just 3% of the reserve baskets of those central banks who deign to disclose their reserve weightings. Sterling, on the other hand, has a 4% weighting (and nearly 6% amongst emerging economy central banks.) If that’s not the carry trade, what is?
Ultimately, the yen should and probably will remain weak on a broad basis until one of two things happens.
If Japan careens back into deep recession, Mrs. Kobayashi will likely curtail her purchases of foreign securities, eliminating a substantial headwind for the yen. At that point, risks of a short-covering rally would increase.
On the other hand, Voldemort and his cronies could finally alter their behaviour and start buying yen like they do euros. If and when that happens (and it might not be that far off), the yen will rise very substantially, and it will be a great trade to be long yen against everything.
For the time being, however, neither of these factors is in play. And while the good Dr. Roubini jets around the world (presumably on someone else’s dime) wringing his hands with worry, those of us who have real P/Ls will remain more concerned with finding out the real story than the interesting story. And if that makes us fools, well, then so be it. I’d rather be a fool than a big fat idiot any day.
25 comments
Click here for commentsSept 5, 1998 USD/JPY 135.90
ReplySept 8, 1998 USD/JPY 112.50
Sept 7 1998 Hardly any trade between the USD/JPY 130 NY close and 121 Mid-session Tokyo & NY levels.
Shit happens. Only fools are unprepared, or worse, on the wrong side of the higher moments for the sake of a coupla' pips-'a-day.
Oh, and if one has a family, and one is reasonably geared into the trade, it would be wise to insure one's paid off one's mortgage and set-aside some mad-money in a secure trust. Creditors can be s right bitch when they're hung out to dry!
I was there, hammer.
Reply1998: 2007 : Dan Quayle : Jack Kennedy. Superficial similarities, but different where it mattered.
Any trade that produces so much unwarranted moralizing has got to be the right thing to do...
If John Calvin were alive today, you could bet your boots he'd be short USD/JPY (and EUR/CHF, naturally.)
Roubini's merry band of Tin-Foil Hat nitwits
Replyhe doesn't understand the trade , doesn't understand that there's someone on the other side of it and doesn't understand the margin involved
other than that , he's spot on
Brad Setser appears to be both open-minded and civil, so I should probably retract any allusions to "all and sundry" at RGE.
ReplyNevertheless, I do think his blog discussion is missing the boat by confusing gross outstanding FX forwards with net outstanding FX forwards, a very considerable difference.
the classroom guys rarely understand investing in the "real" world , understanding relationships between various financial instruments and their responses to macro events , and understanding how different the "college books" are from actually "trading your own book "
ReplyMacroman man said:
Reply"Any trade that produces so much unwarranted moralizing has got to be the right thing to do..."
Any trade that has everyone speculatively stacked up on one side, where such leveraged open interests are wrong-way to the people that can change the rules and where such people have the intensity of interest to do so, face kurtosis up the ass. It's like GKOs....nice yield and makes you look a bona-fide genius until you wake up one day and your USD are to be repaid in Rubbles. But of course all the smart guys think they will be able to unwind before Milton's notorious Pandaemonium.
Sigh. Contrary to popular belief, the entirety of the speculative positioning is NOT short yen, as the Russell Mellon chart clearly shows. Data from Merrill Lynch corroborates this view.
ReplyMoreover, the two sets of people who could fundamentally change the direction of the yen, the Ministry of Finance and the FX Reserve Managers, have demonstrated very little "intensity of interest" to alter the direction of the yen. The former because, well, they are still trying to beat deflation and have an election to win, and the ;atter because the yen doesn't pay them a rewarding rate of return.
One can only presume that Cassandra has been out of US equities since October 19, 1987 (if not October 24, 1929), since it appears that in her/his world, what happened before must necessarily happen again, regardless of circumstance.
I actually like your little tactical portfolio irrespective of your ignoble inferences of my own equity exposure(s).
ReplyBut however one slices it, those long USD/YEN are long of positively skewed but small returns and short the tail. You can rationalize it all you want, and you may - ex-post - be proven correct in your assessment and worthy in regards of your luck. But one should be honest with oneself about where the tail-risk lies. This is rule#1 in avoiding the risk of ruin. Disparaging those that highlight the location of such risks (Setser, Roubini, and you could say the same for Rogoff, Roach, Krugman, Volcker, Issing, Blinder Stig, Rato, etc.) won't make it go away. Good Luck!
the Yen is overvalued and should be shorted
ReplyCassandra
ReplyBelieve me, I am amply aware of the skewed distribution inherent in carry trade returns. Indeed, I have spent most of the last several weeks vascillating and rationalizing not slapping on the full G10 carry trade. If you peruse the site archives from mid-late Jan, you'll find a new excuse every day for delaying the implementation. I still cannot bring myself to buy the GBP and NZD portions of the 'beta plus' portfolio, such is my concern that they are destined for the rubbish heap.
In general, I have no wish to be antagonistic and much prefer civil debate. The style of yesterday's post was a dig that I couldn't resist at a commentator who displays a rather sophomoric penchant for ridiculing those who disagree with him. As I posted today, I have no beef whatsoever with Setser, who appears to conduct himself with the decorum appropriate for a public commentator.
Cheers
Ok, how about this: correlation between the VIX & the US/JPY, and also between VIX and SP500. How do you explain how those graphs looked like during 2006, when QE was ended and assets went down hard with apparent Yen covering, and then back to normal? The only non-carry-trade explanation would be that people invested specifically in JPY instead of risky assets... which is far-fetched. Isn't that a sign that there's (or, for now, let's just say, there WAS) a Yen carry trade?
ReplyGood question, as this is a common fallacy in the marketplace.
ReplyAt the time, long yen WAS the risk-loving position. So was long Swiss. And long BRL. And short kiwi. And short Turkey. And short Aussie. And short dollars.
What do these things have in common? The longs are current account surplus currencies, and the shorts current account deficit currencies.
The thematic trade last April-May was NOT the FX carry trade, but rather the FX rebalancing trade. This was initially kindled by the Iceland downgrade in February, and really got going after the G7 meeting on April 22-23 appeared to create a framework for addressing external imbalances. Cue mad rush for current account surplus currencies; hell, even Macro Man was long yen last May.
However, the current account trade was de-railed by a bout of risk aversion caused not by yen strength but by the Bernanke Fed turning overtly hawkish to regain credibility after the late April 'we may pause' speech and Bartiromo-gate.
The peak in the SPX was May 11, the date the Fed surprised markets by signalling that another rate hike was in the offing, contrary to expectations.
Observe that inflation breakevens rose from 4.52 at the end of March to 4.73 as the Fed was meeting in May. The hawkish statement marked the top in both breakevens and the SPX.
The demise of the popular currency trades soon followed, and USD/JPY bottomed on May 16 at 109. By mid-July, when the SPX was 1235, USD/JPY was back up at 117.50.
This weekend's G7 meeting is unlikely to have a similar impact, as last April appeared to involve the IMF in a mulitlateral framework to smooth exchange rate adjustments. This weekend's meeting is just a run-of-the-mill meeting that will feature a lot of moaning, but in my judgement little of substance. However, the threat of polciy changes arising out of G7 is one of the reasons why some yen shorts are closing and other specs are going long yen into the meeting.
Ok so let me get this. You're saying:
Reply1) The "long deficit country" trade started unwinding in February, not april.
2) this reversal lasted only a couple months, because of fed strength. (i.e. deficit countries are ready to defend their currency).
3) the SPX was down during those couple of month precisely because of fed actions.
4) but since july, the fed has paused, so the SPX rallied...
ok. but how about this:
- how do we explain how commodities & metals went up until late april and then down? somebody supported, and then destroyed, this trade. You say it's the Fed.
Here's the chart:
http://www.minyanville.com/articles/index.php?a=11492
While you are right that during Jan-April the Yen and VIX moved in opposing directions a(and even in other periods) you will notice that the overall graph is one of a very good correlation 8albeit* with a lag (VIX lags Yen). As a matter of fact, even the Jan-April period fits into this lag picture.
So maybe the Fed provided the matches but the Yen provided the fuel. Also, you have something else going against your view: an overwhelming agreement, even among people who disagree with each other, that the Fed is not as powerful as you make it to be.
what do you say?
P.S. in items 2 and 3, when I say "couple months I may May-June, not March-April. Sorry for the confusion.
ReplyYour summary is somewhat incorrect. My version of events is that the long risky carry currency trade started to go awry in February. This led to sell-offs in ISK, TRY, NZD, HUF. USD/JPY traded on the back foot but showed just enough life to keep longs in the trade. Overall, however, carry strategies remained the "risk trade". Chart the Barclay Currency Trader Index (BTOPVAMI index on Bloomberg, if you have one); currency trader performance was negative in Feb and March during this unwind.
ReplyApril G7/IMF announcement changes the environment, and the market switches gears entirely. USD/JPY gaps through erstwhile support, longs stop out and eventually go short, which proves to be a successful trade for three weeks.
You can observe this either through the IMM data, in which even CTAs went long yen (albeit largely as a function of price), or the Barclay index cited above, which rose very, very sharply from late April.
To reiterate: the market lost money as carry strategies underperformed from Feb to mid-April, then made very substantial amounts of money in late April and early May.
Cue the Fed action described above, in which all risky trades lost money, whether it was long gold, short USD/JPY, long stocks, or short USD/BRL.
I am not saying that the Fed attempted to derail a series of speculative bubbles. I am saying that the Fed's actions caused financial conditions to tighten substantially; it was this tightening of financial conditions which killed all the trades that made money in the three weeks before May 11.
To reiterate: the market lost money as carry strategies underperformed from Feb to mid-April, then made very substantial amounts of money in late April and early May.
ReplyOk in this sentence you use 'carry' as reference to pure currency carry trading of deficit countries. I see.
But nevertheless during this time the carry trading of gold and oil and stocks and other risk assets was still ongoing. I understand that you agree there was *some* type of carry trading of all risk assets, but you just don't think it's so much the Yen, right?
Finally, you did not address the Yen/Vix chart. Your initial comments about it made sense, but now that I've noticed they correlate quite amazingly with a lag, I am pretty convinced of either of the following 2 conclusions: (i) there IS a big Yen carry trade and it's a main factor, or (ii) the Yen happens to be a safe haven that people run away to when they see trouble in risky assets. Which of the two...?
What you are calling 'carry' strategies, I would call 'risk' strategies.
ReplyFrom my perspective, long gold can never, ever, ever be a carry trade, as it costs actual money to be long gold (via the cost of capital contango in the futures market, via the cost of storing and insuring in the physical market), whereas whatever currency you have spent to buy gold offers some non-negative rate of return. Long gold is, in my view, the antithesis of the classic carry trade, which relies on interest rate parity not holding.
In late April-early May, the funding currency of choice was USD, for reasons already identified.
As for the VIX...well, I think it is a fairly well-known phenomenon that the VIX is really just the inverse of the SPX. When equities go up, VIX goes/stays down. When equities go down, VIX goes/stays up.
In my worldview, it is entirely consistent that VIX and USD/JPY spiked within a couple of days of each other, and that equities and gold topped within a few days of each other.
As I demonstrate in today's post 'Yen again', claiming that economic losses on the yen somehow foreshadowed economic losses on equities, gold, etc. requires one to explain exactly how currency funds made so much money as USD/JPY went down, and why they started losing money at exactly the same time that USD/JPY bottomed, the SPX peaked, etc.
My answer, the Occam's razor answer, is the simplest answer: the market was long yen in April and May.
"Another way to gauge positioning is to look at Japanese banks’ yen lending to foreigners. Currency traders must ultimately borrow yen when they sell it short, and this borrowing generally filters bank into Japanese banks, with whom other banks have their yen nostros. We can therefore get a read on yen borrowing by looking at Japanese banks’ yen lending."
ReplyCould you please elaborate on how exactly speculative short yen positions in the FX forward and swap market would show up in the cross-border yen lending activity of Japanese banks? Thank you.
Obviously yen borrowing in forward/swap markets doesn't require any physical delivery, and thus does not require physical borrowing.
ReplyHowever, my inquiries to a number of prime brokerage banks revealed that a significant portion of their hedge fund client base are trading spot. Frankly, I found this surprising, as I would have thought it easier to trade to a single forward date and then roll the net exposure a few days before maturity.
Nevertheless, if what I was told is correct (and I have no reason to doubt it), this would require borrowing yen to deliver, which would ultimately feed back into onshore borrowing from a Japanese bank, presumably through the PB's onshore nostro account.
Moreover, yen lending to foreigners reflects another type of 'carry trade' that seems to generate a substantial amount of worrying: yen borrowing by "real" economic agents because the interest rate is lower.
This could represent corporate borrowing from banks or, as is often the case, yen mortgages in places like Korea or Australia.
Certainly the Hungarians have quite a penchant for Swiss franc mortgages; despite the weakness of the Swiss meanwhile, inflation continues to undershoot the bare minimum required for the SNB to pull the trigger on tightening.
Dear Macro Man,
ReplyI enjoyed your comments on this whole 'yen carry trade' business and learned something new... Thank you.
But with regard to to your latest reply on how FFX swaps show up in cross-border lending data:
if most of your anecdotal evidence suggests that hedge funds are doing spot trades, then my question is - are they really trading to earn carry? If they don't borrow yen physically, then they just take directional positions... And if they do borrow yen physically, then where would it appear - on their books at prime brokers? If that's the case, how does this work? Do prime brokers go to a Japanese bank and borrow yen from them using up their own credit limit?.. I'm just wondering if putting more borrowing / lending on their balances is optimal for the prime broker banks, in terms of regulatory capital and such... Woudl be grateful for some more clarification... thanks!
To be honest with you, your query re: prime brokerage banks and their credit exposure exceeds my knowledge of the subject. So the honest answer is: I don't know. Perhaps any readers with a more intimate knowledge of the subject could chime in.
ReplySimilarly, re: trading spot...I was frankly surprised to hear that. I trade to a single date and then roll it to crystallize the gain/loss, and would have assumed that funds would do similarly. Again, anyone at a fund that trades spot, feel free to chime in with an explanation!
As for betting on direction and not just carry...I think that is a very important distinction between the 2000's and the 1990's.
The 1990's were the decade of the pure carry trade: you'd have a pegged exchange rate (whether the ITL, THB, or BRL) that offered a higher interest rate than the anchor currency. However, these trades sowed the seeds of their own destruction. The carry currencies inevitably had something 'wrong' with them: too much inflation, a high current account deficit, or both.
Thus, maintaining the peg implied a consistent appreication of the real exchange rate, which would ultimately prove unsustainable. Volia! ERM crisis-> Tequila crisis -> Asian crisis -> Latam crisis.
Today, our carry trades generally involve free-floating currencies. Obviously, the yen carry trade involved a free-floating currency in 1998, but as I hope I have demonstrated this week, 2007 is almost nothing like 1998 in terms of its magnitude. In any event, playing carry via free floaters offers both the opportunity and threat of capital gains/losses on top of the coupon clipping.
Macro Man,
ReplyInteresting food for thought, presented in a mostly civil manner. Thank you.
What is the source for your data on Japanese bank lending to foreigners? BOJ or BIS or ... ?
The literal answer is that I get it from a data provider, the slightly less literal answer is that it is in the Japanese flow of funds data, and the answer that you are after is that it is compiled by the BOJ, I think.
ReplyHi Macro Man,
ReplyWho turned out to be right, you our Roubini...?
/N.
Well, when it comes to the specific question of whether the yen carry trade would bring down the world economy, I think I was right. It turned out to be a) a complete sideshow, and b) offered plenty of time to get out.
Reply