Yesterday's post generated quite a bit of repartee in the comments section, including a good question from TMM alumnus Polemic about the level of leverage in the financial system. While it certainly seems fair to suggest that leverage in the banking system is lower than it was pre-crisis, courtesy of various regulatory bogeys and the collapse in the velocity of collateral, this doesn't say much about the leverage taken by people who are buying and selling securities.
Although there is no perfect measure of leverage, a simple and widely-cited proxy is the level of margin debt on the NYSE, which has the benefit of a relatively long data history. Macro Man hadn't really looked at this data in a while, since the last time he had it appeared to be relatively benign. So he pulled it up, and lo and behold! It has now reached all-time highs in nominal terms.
Of course, both the size of the economy and the size of the stock market's capitalization have changed as well, so looking at a nominal margin debt figure in isolation may not be particularly useful. Macro Man could only source monthly stock market capitalization data going back to 2003 (if anyone has a longer dataset, feel free to share it!) What we can observe is that while debt as a % of market cap is above the levels prevailing before the crisis, it is not at the threshold that precipitated financial meltdown.
Of course, it's not exactly an apples-to-apples comparison. Referencing NYSE margin debt to total market cap could potentially miss shifts in the relative capitalization of NYSE stocks to those listed on other exchanges. Moreover, it sure would be nice to see what the indicator looked like during the tech bubble in the late 90's.
(Edit: Thanks to SP for sending an alternative datasource that goes back to 1980. As you can see, margin seems to have continued its inexorable trend drift higher of the last two decades, though without the parabolic rise that has augured previous secular tops.)
One thing we can look at, however, is the ratio of margin debt to GDP. Again, it's not a perfect indicator, but at least it has the benefit of history that predates even the 1980's bull market. So what happens when we pull this data up?
Zowie! The current 2.5% level represented the top of previous leverage cycles, peaking there in April 2000 (the month after the Nasdaq peak) and July 2007 (the month the Bear credit funds blew up, ushering in the financial crisis that still resonates today.)
Now, as some readers will no doubt point out, those previous peaks in market leverage were accompanied by peaks in real-economy leverage that exacerbated the impact of a market down-turn. Point taken, but the aphorism that the "market is not the economy" cuts both ways. If leverage and pricing can build up in the absence of brisk economic growth, presumably they can draw down without the necessity of a recession.
Now, while the third chart above would certainly suggest that a market top could be near, Macro Man is not going to stand on a soapbox fulminating that the End Is Nigh and that OMG! #Stocksovervaluedinnit. He would, however, suggest that the data would appear to support his notion that despite all the warnings, markets may be ill-equipped to deal with an eventual liquidity withdrawal. While such a view may be consensus, it would not appear to be positioned for in certain market segments.
And if it's not positioned for, how consensus can it really be?
ED. NOTE: Sadly, Macro Man has seen his Bloomberg transition account withdrawn this week. While he can still of course get news and prices, data and charts will be somewhat harder to come by. As such, it will be nigh-on impossible to provide quality insight on a daily basis. (Some readers may wish to point out that that hurdle had already proven to be insurmountable for your author.)
As such, updates are likely to be a little more sporadic and long-form in nature, perhaps not unlike those from TMM when they were in their transition phase. Macro Man already has one in mind that he hopes to finish by the end of the week.
Although there is no perfect measure of leverage, a simple and widely-cited proxy is the level of margin debt on the NYSE, which has the benefit of a relatively long data history. Macro Man hadn't really looked at this data in a while, since the last time he had it appeared to be relatively benign. So he pulled it up, and lo and behold! It has now reached all-time highs in nominal terms.
Of course, both the size of the economy and the size of the stock market's capitalization have changed as well, so looking at a nominal margin debt figure in isolation may not be particularly useful. Macro Man could only source monthly stock market capitalization data going back to 2003 (if anyone has a longer dataset, feel free to share it!) What we can observe is that while debt as a % of market cap is above the levels prevailing before the crisis, it is not at the threshold that precipitated financial meltdown.
Of course, it's not exactly an apples-to-apples comparison. Referencing NYSE margin debt to total market cap could potentially miss shifts in the relative capitalization of NYSE stocks to those listed on other exchanges. Moreover, it sure would be nice to see what the indicator looked like during the tech bubble in the late 90's.
(Edit: Thanks to SP for sending an alternative datasource that goes back to 1980. As you can see, margin seems to have continued its inexorable trend drift higher of the last two decades, though without the parabolic rise that has augured previous secular tops.)
One thing we can look at, however, is the ratio of margin debt to GDP. Again, it's not a perfect indicator, but at least it has the benefit of history that predates even the 1980's bull market. So what happens when we pull this data up?
Zowie! The current 2.5% level represented the top of previous leverage cycles, peaking there in April 2000 (the month after the Nasdaq peak) and July 2007 (the month the Bear credit funds blew up, ushering in the financial crisis that still resonates today.)
Now, as some readers will no doubt point out, those previous peaks in market leverage were accompanied by peaks in real-economy leverage that exacerbated the impact of a market down-turn. Point taken, but the aphorism that the "market is not the economy" cuts both ways. If leverage and pricing can build up in the absence of brisk economic growth, presumably they can draw down without the necessity of a recession.
Now, while the third chart above would certainly suggest that a market top could be near, Macro Man is not going to stand on a soapbox fulminating that the End Is Nigh and that OMG! #Stocksovervaluedinnit. He would, however, suggest that the data would appear to support his notion that despite all the warnings, markets may be ill-equipped to deal with an eventual liquidity withdrawal. While such a view may be consensus, it would not appear to be positioned for in certain market segments.
And if it's not positioned for, how consensus can it really be?
ED. NOTE: Sadly, Macro Man has seen his Bloomberg transition account withdrawn this week. While he can still of course get news and prices, data and charts will be somewhat harder to come by. As such, it will be nigh-on impossible to provide quality insight on a daily basis. (Some readers may wish to point out that that hurdle had already proven to be insurmountable for your author.)
As such, updates are likely to be a little more sporadic and long-form in nature, perhaps not unlike those from TMM when they were in their transition phase. Macro Man already has one in mind that he hopes to finish by the end of the week.
9 comments
Click here for commentsNYSE margin debt might be a decent proxy but that's it. For example books I'm familiar with are all on swap, so despite a fair amount of leverage it's all unrecorded by that measure.
ReplyI think it's fair to say that swap leverage is probably quite a bit lower than in the past, given the collapse in the velocity of collateral (ie decline in rehypotehcation post Lehman) and the regulatory burden of moving things on exchange, etc.
ReplyThen there are issues like 'hidden leverage', wherein ETFs have granted access and 'liquidity' in markets that were previously closed to non specialists. How those markets perform and react when investors wish to sell en masse remains to be seen.
MM, the Fed websites have great wealth of data and some easy to use charts as well.
ReplyLooking at overall Fund Flows credit relative to nominal GDP, you can see we have stabilized at about 3.5x, though thats up quite a lot from the 90s and early part of last decade but off from 3.8 or so in 2008
http://imgur.com/CkJBWRv
ReplyCredit / GDP
Dude, just use FRED2. Total credit to US private non-financial sector as percentage of potential GDP is at http://research.stlouisfed.org/fred2/graph/?g=GAY
ReplyThe only big peaks on the chart are the housing peaks in 1989 and 2007.
MM, do you not fancy forking out £1600 pcm for a Bloomy, PA?
ReplyThanks me old mucker for that post.
ReplyAs for Bloomberg, the loss of it after 20 years of use made me feel i had nicotine withdrawal after a 60 a day habit. The prices they charge are pretty similar to a serous addiction habit too.
For data, try http://www.quandl.com/.
ReplyNo affiliation with the site but you might find it useful.
I have the impression that leverage in equities is almost meaningless compared to fixed income these days. With the risk parity fad, leveraged bond portfolios have even become mainstream for pension funds. But there's no data.
Reply