View Full Version : Deflationistas get the facts wrong about Japan
http://www.itulip.com/images/humptydumbass.jpgDeflationista get the facts wrong about Japan
Japan's post bubble economy is held up by many economic commentators as a model of America's deflationary future. Aside from major differences between the US and Japanese economies, the assertion that Japan has experienced a continuous deflation does not hold up to examination of the evidence. Looking at the actual changes in the money supply, inflation, and economic growth after application of monetary and fiscal injections, rather than only whether CPI or GDP were positive in Japan, is key to understanding what happened in Japan over the past 18 years since the stock market and property bubbles collapsed there. In fact, the response of the economy to stimulus offers a far more relevant lesson for anyone looking to Japan for lessons on deflation and inflation.
We start by examining the 1996 to 1997 period. The CPI went up 5% following stimulus applied starting in 1993, and 4% in the 1999 to 2000 period after fiscal stimulus kicked in that was applied years before. GDP grew all the way from -4% in early 2003 to spike to over 10% after massive stimulus in 2001 and 2002. While the monetary base (M3) didn't go above 10% growth rate for any significant length of time until 2001, the Nikkei well over doubled after that M3 pumping as well as responding strongly to other base expansions.
http://www.nowandfutures.com/images/boj_money_key_stats1990on.png
http://www.nowandfutures.com/images/boj_money_key_stats1990on2.png
Next we consider what money supply, credit, and government debt were doing in the 1980s in order to push real estate and the Nikkei, etc. to such heights (asset price inflation); they averaged about 14% and were relatively even in their growth rate -- and were 20-25% in the early 1970s. Both base and government debt have gone up and down like a yo-yo since 1990.
Then we have the cultural differences. As iTulip has noted for years, where The Great Depression left an indelible fear of goods price deflation on the population and ppolicy makers, the Japanese fear the opposite due to the hyperinflation that wiped out what wealth remained after WWII. That affected the behavior of both the BoJ and MoF, which therefore affected velocity and inflationary expectations. Credit growth can't and won't change when the population of a country expects lower prices ahead.
Stated even more simply, money, credit and debt growing at around 14% created lots of asset price inflation and GDP growth in the 1980s, yet why is anyone surprised when money, credit and debt growing erratically and at around 7% average didn't reflate asset prices and grow GDP since 1990?
The more recent, the Japan CPI jump to around 8% in my opinion had more to do with global factors than local ones -- the oil and commodity price inflation primarily. But up to 2004 to 2005, the effects from the 2001 to 2002 pumping are quite clear when normal monetary flags are taken into account. Not only commodity prices by asset prices went up; the Nikkei and real estate prices responded.
The charts above are intended to show that inflation can be created by a government intent on creating it, and also to posit that the 10%+ metric as a rough guideline, and also to educate some of the deflationistas who continually and egregiously get the actual facts about Japan very, very wrong.
Drawing conclusions about deflation from the Japanese experience while ignoring the men behind the curtain and the actual facts can be dangerous.
See also: Anatomy of a credit crunch induced bankruptcy (http://www.itulip.com/forums/showthread.php?p=72157#post72157) - Why debt deflation is not the same as goods price deflation
iTulip Select (http://www.itulip.com/forums/showthread.php?t=1032): The Investment Thesis for the Next Cycle™
__________________________________________________
To receive the iTulip Newsletter or iTulip Alerts, Join our FREE Email Mailing List (http://ui.constantcontact.com/d.jsp?m=1101238839116&p=oi)
Copyright © iTulip, Inc. 1998 - 2009 All Rights Reserved
All information provided "as is" for informational purposes only, not intended for trading purposes or advice. Nothing appearing on this website should be considered a recommendation to buy or to sell any security or related financial instrument. iTulip, Inc. is not liable for any informational errors, incompleteness, or delays, or for any actions taken in reliance on information contained herein. Full Disclaimer (http://www.itulip.com/GeneralDisclaimer.htm)
Edited and re-posted to elicit comment.
rdgmail
01-24-09, 03:24 PM
Bart:
Very much much appreciate your data and analysis as always. I don't question your conclusions, but your conclusions do raise some follow-up questions and comments:
Itulip has been emphatic that deflationists should distinguish between all goods price deflation (or inflation) versus asset price deflation. I agree that's an important distinction. In that sense, could the deflationists you're referring to be correct, i.e., perhaps they focus on the Nikkei and Japan real estate prices and other assets, which are nominally waaay down from their peaks in 1989? When I think about Japan's Lost Decade, I think of the asset price deflation, not all goods prices, gdp or anything else.
Similarly in the US today, I presume people will not feel like we've "recovered" from this depression until their houses are worth what they were in August 2005 and their stocks are worth what they were in October 2007 (or 2000 if your prefer). Of course this won't help the millions who've already lost their houses and liquidated their stocks. But if the test for a government (or Central Bank) is its ability to reflate asset prices instead of all goods prices or even GDP, what lessons can the U.S. learn from Japan and it's late-to-the-game quantitative easing in 2001?
How much printing/ devaluing of currency would the U.S. Treasury and Fed have to do to get "asset inflation"? And of course, because all goods prices have not and will not fall like asset prices, how much all goods inflation will result if the Fed/ Treasury is successful in reflating asset prices to their prior lofty highs (obviously talking in nominal terms here)?
EasternBelle
01-25-09, 01:11 AM
Sorry, but I can't see a thing in those graphs, they wiggle around. I can't see time lags or correlations. So bart's conclusions evade me.
Eastern Belle
http://www.itulip.com/images/humptydumbass.jpgDeflationista get the facts wrong about Japan
Credit growth can't and won't change when the population of a country expects lower prices ahead.
The charts above are intended to show that inflation can be created by a government intent on creating it, and also to posit that the 10%+ metric as a rough guideline, and also to educate some of the deflationistas who continually and egregiously get the actual facts about Japan very, very wrong.
__________________________________________________
To receive the iTulip Newsletter or iTulip Alerts, Join our FREE Email Mailing List (http://ui.constantcontact.com/d.jsp?m=1101238839116&p=oi)
Copyright © iTulip, Inc. 1998 - 2009 All Rights Reserved
All information provided "as is" for informational purposes only, not intended for trading purposes or advice. Nothing appearing on this website should be considered a recommendation to buy or to sell any security or related financial instrument. iTulip, Inc. is not liable for any informational errors, incompleteness, or delays, or for any actions taken in reliance on information contained herein. Full Disclaimer (http://www.itulip.com/GeneralDisclaimer.htm)
Okay, I'll bite. Is that not contradictory. (No credit growth if people feel prices will fall... A determined Gov/CB can allways produce inflation?)
Please, could you explain how the apparent contradiction above resolves itself and fits into the big picture macro economic puzzle for us "simple simon" types.
Thanks!
V/R
Simple Simon ;)
Bart:
Very much much appreciate your data and analysis as always. I don't question your conclusions, but your conclusions do raise some follow-up questions and comments:
Itulip has been emphatic that deflationists should distinguish between all goods price deflation (or inflation) versus asset price deflation. I agree that's an important distinction. In that sense, could the deflationists you're referring to be correct, i.e., perhaps they focus on the Nikkei and Japan real estate prices and other assets, which are nominally waaay down from their peaks in 1989? When I think about Japan's Lost Decade, I think of the asset price deflation, not all goods prices, gdp or anything else.
Similarly in the US today, I presume people will not feel like we've "recovered" from this depression until their houses are worth what they were in August 2005 and their stocks are worth what they were in October 2007 (or 2000 if your prefer). Of course this won't help the millions who've already lost their houses and liquidated their stocks. But if the test for a government (or Central Bank) is its ability to reflate asset prices instead of all goods prices or even GDP, what lessons can the U.S. learn from Japan and it's late-to-the-game quantitative easing in 2001?
How much printing/ devaluing of currency would the U.S. Treasury and Fed have to do to get "asset inflation"? And of course, because all goods prices have not and will not fall like asset prices, how much all goods inflation will result if the Fed/ Treasury is successful in reflating asset prices to their prior lofty highs (obviously talking in nominal terms here)?
I very much agree that its important to distinguish between goods and asset inflation, and it even goes further than that in the sense that paper assets like stocks or housing are under performing hard assets like precious metals.
But my real point about Japan since the '90s (which it appears that you understand if I'm reading you right) is that the deflationistas have gotten it wrong in at least three areas:
1. When true total money supply (M3, credit, gov't debt, etc. as opposed to the Austrian TMS) grows fast enough, inflation does and did result.
2. Cultural values do make a huge difference, as in the fear of inflation in Japan vs. the fear of deflation in the US.
3. If the money creation machine goes fast enough, even assets like real estate do eventually respond.
My basic point was also covered in Stated even more simply, money, credit and debt growing at around 14% created lots of asset price inflation and GDP growth in the 1980s, yet why is anyone surprised when money, credit and debt growing erratically and at around 7% average didn't reflate asset prices and grow GDP since 1990?
Your point about the average person not believing its over in the US until "until their houses are worth what they were in August 2005 and their stocks are worth what they were in October 2007" is also quite valid and is also somewhat echoed by their experiences & opinions in the '70s and early '80s.
The bottom line and way overly simplified is that the lesson of Japan is twofold:
1. Fighting the trends and hard vs. paper asset cycles (or KaPoom) etc. is very expensive.
2. If one wants to get them to recover to previous nominal values (regardless of real values), lots of money creation is required.
And in this case and to your "how much printing/ devaluing" is required, my minimum 10%+ even growth rate (plus normal monetary lags) is my best guess for what it takes. The consistent, even, high growth rate is what it takes to make velocity go up too... velocity at its simplest just measures inflationary expectations.
And of course it depends on which assets one wants to inflate - it will take a lot more money creation to boost housing prices than precious metal or other hard asset prices. In Japan's case, it took 15-20%+ growth rate to have an effect on Tokyo real estate prices for example.
And here's hoping that I did address your questions & comments well enough.
Sorry, but I can't see a thing in those graphs, they wiggle around. I can't see time lags or correlations. So bart's conclusions evade me.
Eastern Belle
Would that I was a whole lot better at presenting data & conclusions in many of my charts, and I do understand - you're far from the first person to have trouble with them. I frankly am at a loss in knowing how to make them better.
One suggestion that may help - look at the Nikkei data in the second chart and notice how it does respond to base going up in the same chart. Then notice at the points where it does respond more than expected that debt (from the first chart) is also going up fast.
Perhaps I should make another chart that adds base and gov't debt growth rates to make the correlation clearer...
metalman
01-25-09, 11:31 AM
Would that I was a whole lot better at presenting data & conclusions in many of my charts, and I do understand - you're far from the first person to have trouble with them. I frankly am at a loss in knowing how to make them better.
One suggestion that may help - look at the Nikkei data in the second chart and notice how it does respond to base going up in the same chart. Then notice at the points where it does respond more than expected that debt (from the first chart) is also going up fast.
Perhaps I should make another chart that adds base and gov't debt growth rates to make the correlation clearer...
why not dump the data into excel and make charts that are more readable? your excellent analysis is lost in the squiggly charts.
Credit growth can't and won't change when the population of a country expects lower prices ahead.
The charts above are intended to show that inflation can be created by a government intent on creating it, and also to posit that the 10%+ metric as a rough guideline, and also to educate some of the deflationistas who continually and egregiously get the actual facts about Japan very, very wrong.
Okay, I'll bite. Is that not contradictory. (No credit growth if people feel prices will fall... A determined Gov/CB can allways produce inflation?)
Please, could you explain how the apparent contradiction above resolves itself and fits into the big picture macro economic puzzle for us "simple simon" types.
Thanks!
V/R
Simple Simon ;)
The simplest answer to the apparent contradiction is that credit is only one component of total money supply. Total money supply in my opinion is credit plus M3 plus gov't debt, plus elements of derivatives etc.
The key in my opinion is starting with one of the absolute basic definitions of money - a medium of exchange - rather than getting excessively focused on one element like M3 or credit. Credit is accepted as a medium of exchange, as are all the rest - M3, gov't debt and even derivatives and other items. And at the risk of getting too far from your basic question, things like debt deflation also have a huge effect on total money supply.
The deflationistas consistently get it wrong by ignoring that credit is only one part of total money supply, and by having an incomplete definition of total money supply. Large increases in gov't debt or large increases in monetary base do cause total money supply to grow, and inflation (whether relative or not) to occur.
Here's another picture, this time from the US during the 1920-1940 period, that may show it more clearly.
http://www.nowandfutures.com/download/m1m3_gdp_debt_1920_1940.png
Notice the increased growth rate in both base (light blue line) and gov't debt (burgundy line) in 1930 and 1931... and that GDP responded after time lag. The chart is overly simplistic since there are many factors not shown like velocity, CPI or the dollar devaluation in 1933... but I believe that the evidence is clear, especially given that when both peaked in the 1935-36 time frame, the recession of 1937-38 resulted.
This chart is overly busy for the purpose, but it covers the same period... and shows that credit growth (the dotted line) lagged the growth in base, gov't debt and even M3 by quite a bit. Credit growth is a lagging component of money supply growth during severe disinflationary or deflationary periods.
http://www.nowandfutures.com/download/m1m3_1920-1940.png
Did that answer your question and address your concern?
why not dump the data into excel and make charts that are more readable? your excellent analysis is lost in the squiggly charts.
The huge majority of all my charts and data are already in Excel. I probably have one of the largest Excel workbooks ever - currently its about 1.1GB in memory.
If you have an idea on how to make those less spaghetti like and more communicative, I sure am listening. The primary reason I created two charts is to avoid at least some of their spaghetti nature. Trying to get across semi-complex relationships is a b*tch... even using two scales seems to confuse more than a few folk. *sigh*
In one sense, I'm damned if I do and damned if I don't too - since if I present only one or two data series on one chart, someone will inevitably wonder why I didn't also include some other data series.
metalman
01-25-09, 01:46 PM
The huge majority of all my charts and data are already in Excel. I probably have one of the largest Excel workbooks ever - currently its about 1.1GB in memory.
If you have an idea on how to make those less spaghetti like and more communicative, I sure am listening. The primary reason I created two charts is to avoid at least some of their spaghetti nature. Trying to get across semi-complex relationships is a b*tch... even using two scales seems to confuse more than a few folk. *sigh*
In one sense, I'm damned if I do and damned if I don't too - since if I present only one or two data series on one chart, someone will inevitably wonder why I didn't also include some other data series.
why not adopt the itulip excel chart formula... primary colors, white background, legend on top, no more than four items per chart... everyone else has. i get calculated risk newsletter an the chart layout is a near itulip dupe. why not go with what works?
why not adopt the itulip excel chart formula... primary colors, white background, legend on top, no more than four items per chart... everyone else has. i get calculated risk newsletter an the chart layout is a near itulip dupe. why not go with what works?
Besides the huge amount of work to do that to the over 1500 charts I have, I already have white backgrounds and very seldom use more than four items per chart.
I find it difficult to believe that moving the legend to the top and changing to primary colors will help much... but will look at using primary colors more frequently, and did just change those two charts to only primaries.
Anyone - do primary colors make that much difference?
Should I even avoid using a golden color when actually charting gold?
The simplest answer to the apparent contradiction is that credit is only one component of total money supply. Total money supply in my opinion is credit plus M3 plus gov't debt, plus elements of derivatives etc.
Bart, got it. (yes you answered my question)
You and I have the same definition of money supply (I can buy a car with a credit card just as easily as I can with cash or check).
Here is the problem as I see it.
Please note I'm approaching this from a non-idealogical prospective, so just hear me out for a second.
There was at the credit peak something on the order of $1 Quadrillion (+/-) in notional derivatives. Okay does that not dwarf the entire remainder of the money supply by an order of magnitude or two? If that's the case It would seem to take Zimbabwe like inflation to counteract the credit deflation because the other components of the money supply are so small in comparison. So I ask the deflation question in this context: If derivatives are the predominant component in the money supply, can't the CBs inflate with reckless abandon and still not be able to produce a meaningful inflationary jump, because of the size mismatch problem?
I don't have answers, But that question I just asked would seem to be the crux of the whole kitten-kaboodle.
What are your thoughts Bart?
metalman
01-25-09, 02:45 PM
Besides the huge amount of work to do that to the over 1500 charts I have, I already have white backgrounds and very seldom use more than four items per chart.
I find it difficult to believe that moving the legend to the top and changing to primary colors will help much... but will look at using primary colors more frequently, and did just change those two charts to only primaries.
Anyone - do primary colors make that much difference?
Should I even avoid using a golden color when actually charting gold?
send the data from one of your charts to fred and ask him to itulipize it to see if it's worth the trouble?
Bart, got it. (yes you answered my question)
You and I have the same definition of money supply (I can buy a car with a credit card just as easily as I can with cash or check).
Here is the problem as I see it.
Please note I'm approaching this from a non-idealogical prospective, so just hear me out for a second.
There was at the credit peak something on the order of $1 Quadrillion (+/-) in notional derivatives. Okay does that not dwarf the entire remainder of the money supply by an order of magnitude or two? If that's the case It would seem to take Zimbabwe like inflation to counteract the credit deflation because the other components of the money supply are so small in comparison. So I ask the deflation question in this context: If derivatives are the predominant component in the money supply, can't the CBs inflate with reckless abandon and still not be able to produce a meaningful inflationary jump, because of the size mismatch problem?
I don't have answers, But that question I just asked would seem to be the crux of the whole kitten-kaboodle.
What are your thoughts Bart?
Cool - glad it made sense. That total money supply area is in my opinion the biggest problem with the Austrian approach, as well as with many other analysts.
Just to be clear too, I don't have the answers either - those charts are just part of a semi-educated hypothesis. But they also show a huge hole and logic/fact problem in the deflationista viewpoints - and also hopefully that I'm at least somewhat not into ideology vs. actual facts and history. All I really am is just some guy that got ticked off when the Fed discontinued M3 and reconstructed & published it first - and was frankly surprised by the reactions - and have continued as best I know how with exposing other behind the scenes or little known & understood aspects of CBs and governments, and also publishing & highlighting other various poorly known facts. And I'm humble about it too... ;)
I also won't exclude and haven't excluded the deflation or crash & major doom stories in the sense that I'm personally positioned as best I know how to protect myself under any scenario.
In theory of course you're right about derivatives totals against the capabilities of CBs and governments, but I don't believe that the raw derivatives total tells anywhere near the full story.
First, they're primarily a zero sum game and if one nets out all the positions within each individual large participant, the totals go *way* down. That is one possible part of an outcome that's hardly ever mentioned... and the Fed & Treasury & others did strong arm banks to do just that (last October if memory serves). CDS totals dropped something like $25 trillion almost overnight and went into the $30 trillion area from about $57 trillion.
And regarding their zero sum nature - if institution x loses y trillions, then instituion z gains those y trillions. So its literally impossible for the net losses to be anywhere near close to that total.
Second, the majority of the problems are in CDS instruments and they're "only" 6-10% of the total.
As an aside, I've heard folk mention that quadrillion total but have never seen it in actual or documented reports - the last world grand total of derivatives from the BIS was around $684 trillion as of mid 2008.
Third, the huge majority of world derivatives are in the interest rate area - $458 trillion or about 67%, and hedging is so rampant in that area that its my understatement of the year.
Fourth, the entire total of derivatives is not part of the money supply. The actual net total is "only" about $20 trillion without leverage, so although the multi trillions in losses so far are very real, the total losses can't be anywhere near the total - as also noted above in the zero sum point.
Also keep in mind how many trillions of internal derivatives have literally evaporated with banks going under just since last June.
How big the losses are or will be - who knows with any real confidence although I've tried to take a shot at it with my various BKX adjusted charts... but my best guess is that they're smaller than what CBs and governments can deal with... and don't forget their ability to do things behind the scenes or lie etc. Its literally a confidence game, and in both meanings... and vested interests win much more often than they lose.
This is also where things like Robert Mundell's Intor world currency come in, as well as truly ugly things like war or depopulation scenarios.
In other words and to recap, there's a huge amount of "devils in the details".
All we can do is CYA as best we can, look to history for workable parallels, use logical hypotheses to illustrate perceived shortcomings in other and incomplete views, and try our best to divine what's ahead... and of course keep our tinfoil hats firmly in place. ;)
and of course keep our tinfoil hats firmly in place. ;)
I Keep mine under my ball cap. that way it's less noticable in public.
(I wear i for good reason of course)
See
http://www.haarp.net/
I Keep mine under my ball cap. that way it's less noticable in public.
(I wear i for good reason of course)
See
http://www.haarp.net/
That's one of the great things about iTulip... we all harp about this and harp about that... :rolleyes:
And don't tell anybody, but I do the same with a hat... the natives get restless otherwise... ;)
why not adopt the itulip excel chart formula... primary colors, white background, legend on top, no more than four items per chart... everyone else has. i get calculated risk newsletter an the chart layout is a near itulip dupe. why not go with what works?
Thanks bunches for the offline help metalman.
Your points about primary colors and scaling should help my future charts be more useful and understandable. I was too close to the problems to see the simpler answers.
EasternBelle
01-25-09, 10:32 PM
bart,
thanks for trying to clarify your charts.
Looking at them again, if you do not attach meaning to the curves in the sense that you "know" M3, base or credit "should" do something to GDP or the Nikkei, have you tested that you can correlate those series with a reasonable confidence (mathematically speaking)?
Just looking at them I could not see any correlation say between Nikkei and Base, Nikkei and M3, Nikkei and Credit in your yearly growth change chart.
Maybe there is a negative correlation between Base and Credit.
As to the first chart I could maybe see that there was a time lag between GDP growth and big spikes in Base and/or Credit. But in more "normal" growthrates as seen in that 20 year period I could not see anything of a correlation with or without a time lag.
Hope that a untrained eyes views are of help to you.
From a readers perspective it is very confusing to have a left and right hand side of a chart with different "0" lines.
EasternBelle
Bart,
I am a big fan of many of your charts but metalman's comments are on-target.
Most of your graphs are fine but sometimes you put too many (more than 3) series on one graph.
Perhaps the font sizes you use are a little big for my taste.
In any case, fix a house style and stick to it.
Wikipedia have some useful suggestions (http://en.wikipedia.org/wiki/Wikipedia:How_to_create_graphs_for_Wikipedia_artic les) for how to prepare graphs, some of which applies.
In particular I would suggest that Excel is sucky for graphs... You may want to try gnuplot just for plotting, though I understand a move from Excel would be painful. I prepare graphs for publication very often in my day job and use matlab, though that may not be a suitable tool for you.
we_are_toast
01-26-09, 07:15 AM
Thanks for the charts, and the analysis.
And regarding their zero sum nature - if institution x loses y trillions, then instituion z gains those y trillions. So its literally impossible for the net losses to be anywhere near close to that total.
Isn't the problem here that institution x doesn't have the trillions to pay institution z, and z can't pay pension fund C, and C bought the CDS to insure the GM bonds it bought and really needs the money because the CDS C bought from school district B on the Lehman MBS bond has already gone belly up?
If enough links in the chain don't pony up, does it still remain a zero sum game?
If enough links in the chain don't pony up, does it still remain a zero sum game?
Yes. It will still remain a zero sum game, respectively the direct losses (such as bank defaults) cannot be amplified in bigger net losses by derivatives.
bart,
thanks for trying to clarify your charts.
Looking at them again, if you do not attach meaning to the curves in the sense that you "know" M3, base or credit "should" do something to GDP or the Nikkei, have you tested that you can correlate those series with a reasonable confidence (mathematically speaking)?
Just looking at them I could not see any correlation say between Nikkei and Base, Nikkei and M3, Nikkei and Credit in your yearly growth change chart.
Maybe there is a negative correlation between Base and Credit.
Yes I have - the correlation between base plus gov't debt and GDP is about .52 with a time lag of 18-24 months when using the period 1994 to current day.
I also goofed somewhat in including M3 & credit in the charts since they draw away from the main points. I was just trying to get ahead of some folk who I expected to ask - "Yeah, but what about M3 & credit or the Nikkei?", and shot myself in the foot. The answer is that the newest chart below stands on its own and does not need M3 or credit or the Nikkei to show the deflationista shortcomings and massive logic errors.
I may start another thread about the correlations set and the Nikkei too - I believe I can prove the relationship, and with standard math correlations too, even R squared.
As to the first chart I could maybe see that there was a time lag between GDP growth and big spikes in Base and/or Credit. But in more "normal" growthrates as seen in that 20 year period I could not see anything of a correlation with or without a time lag.
And that really is one of the main points I was trying to make, so obviously the primary colors did actually help some.
The real effects on GDP or CPI came only after growth rates in base + debt got high enough... and in plain English, the last paragraph on the new chart tells the rest of the story.
And there are effects on GDP & CPI from things like the Asian crisis, Y2K, etc. too.
Hope that a untrained eyes views are of help to you.
From a readers perspective it is very confusing to have a left and right hand side of a chart with different "0" lines.
EasternBelle
Untrained eyes are indeed quite helpful - its so very easy to get to close to the trees and forget the forest... and I'm far from a pro in chart design, etc, too. Thanks for the feedback - its been quite helpful.
I had thought that moving the monetary base series over to its own axis and showing the full range of data (instead of chopping off the extreme tops and bottoms) would be better than what I had, but it obviously backfired.
Here's a brand new chart (version 6 or so - I've lost count) and I've hopefully made the basic points a whole lot clearer too, both by combining base & gov't debt growth & shrinkage in one line and making it more easy to see the lagged correlations, but have also added the basic logic in a text box on the actual chart itself.
One specific I can point out since its so obvious - base & debt growth (red line) crossed below 5% on the way down in 3/06... and GDP crossed through zero in 3/08, an almost perfect 24 month lag.
A bit of a warning too - the actual monetary lags of base and gov't debt are different and they're also not the same in US & Japan either. I've played a bit fast & loose by using an 18-24 month lag... but it does work reasonably well, especially after taking things like the Asian crisis or various other major issues into account.
http://www.nowandfutures.com/images/boj_money_key_stats1990on3.png
Bart,
I am a big fan of many of your charts but metalman's comments are on-target.
Most of your graphs are fine but sometimes you put too many (more than 3) series on one graph.
Perhaps the font sizes you use are a little big for my taste.
In any case, fix a house style and stick to it.
Wikipedia have some useful suggestions (http://en.wikipedia.org/wiki/Wikipedia:How_to_create_graphs_for_Wikipedia_artic les) for how to prepare graphs, some of which applies.
In particular I would suggest that Excel is sucky for graphs... You may want to try gnuplot just for plotting, though I understand a move from Excel would be painful. I prepare graphs for publication very often in my day job and use matlab, though that may not be a suitable tool for you.
Thanks for the feedback T, good points all.
I will try more to avoid more than 3 or 4 series on one chart in the future, but as you also probably know sometimes folk besides me do actually want to see lots of data on one chart so they don't have to keep flipping from one to another.
I had to chuckle a bit on the font sizes too, since my very first charts had a smaller default font... and I got flack from the 40 & over crowd due to vision issues. I just bring it up to illustrate that I'm damned if I do and damned if I don't sometimes.
I am working towards a house style bit by bit too. I'd mention the name of one of the folk who really liked the various pastel colors and encouraged me to use them, but it would just sound like sour grapes.
I do agree about Excel - it's not a professional charting tool with all the needed flexibility, even in the newest version. I did actually explore some alternatives about 18 months ago, and gnuplot was one of them. Nick at sharelynx.com and I had many good conversations in the area too.
I also have over 20 years of experience as a DBA, Data Architect, *nix jockey and (over priced ;) ) consultant to Fortune 100 companies, so I know how much work it takes to get interfaces reliably going - both internal and external (to other web sites).
I chose to remain with Excel so that the web site, charts and maintenance issues didn't turn into a full time job even though I know its not the best tool. After all, the site is 100% free, I am retired, and do want to have more of a life away from computers and their quirks.
Thanks again for the feedback though.
Isn't the problem here that institution x doesn't have the trillions to pay institution z, and z can't pay pension fund C, and C bought the CDS to insure the GM bonds it bought and really needs the money because the CDS C bought from school district B on the Lehman MBS bond has already gone belly up?
If enough links in the chain don't pony up, does it still remain a zero sum game?
Symbols covered it well and I agree that its still zero sum, and will also add my $.02 worth.
There are indeed a few "institution x" folk who don't have the millions or billions or trillions to pay, and that's also where the Fed & Treasury and Congress (the taxpayer) becomes a backstop. That's also where insurance companies (besides AIG) come in as a backstop.
In other words, *somebody* comes up with the missing billions, and part of it is stockholders equity going to or close to zero.
That's also where arm twisting occurs and why things like the Merrill Lynch merger occured - that way the buyer's & seller's derivatives offset each other and disappear from the existing toxic stock.
The bottom line and bad news is that most banks on the planet are insolvent... and the good news part is that its far from the first time that it has happened. We're still here and managed to get through all previous crises, albeit at very steep prices - not just measured in money.
rdgmail
01-26-09, 09:16 PM
Bart:
Not to throw in too much of a monkey wrench, but have you considered trying to determine whether there is a correlation between Japan's money printing, particularly starting in 2001, and the resulting carry trade effects, if any? Is it possible that Japan's quantitative easing had a bigger effect on inflation and GDP than you've so far measured - not on Japan's economy, but instead on U.S. equities, the mortgage backed security market or other markets outside of Japan? In other words, did Japan export a major portion of the effects of its money supply growth (inflation/ GDP) to other countries?
If so, how might the current U.S. money growth effect other countries' inflation or GDP?
Bart:
Not to throw in too much of a monkey wrench, but have you considered trying to determine whether there is a correlation between Japan's money printing, particularly starting in 2001, and the resulting carry trade effects, if any? Is it possible that Japan's quantitative easing had a bigger effect on inflation and GDP than you've so far measured - not on Japan's economy, but instead on U.S. equities, the mortgage backed security market or other markets outside of Japan? In other words, did Japan export a major portion of the effects of its money supply growth (inflation/ GDP) to other countries?
If so, how might the current U.S. money growth effect other countries' inflation or GDP?
Nothing like simple questions with easy answers, eh? :D
One of the problems is simply finding any data that one can point to and say with any reasonable certainty that "this is because of the carry trade" or "this amount of some international yen flow is part of the carry trade".
You're of course correct that some of the money creation starting in 2001 did export some of the inflation, but something that I've never seen mentioned when that's brought up is that some of the exported inflation also came back to Japan in the form of export business and investing.
But there actually is a fairly simple answer. although it may not be what you expect.
Here's a picture of Japan since 2000:
http://www.nowandfutures.com/images/boj_money_short_term2.png
Here's 70% of the world (Japan, US, Euro area and China):
http://www.nowandfutures.com/images/world_gdp_money2.png
Perhaps I'm being too flippant, but the correlation is pretty clear for me given that broad money does include CDs, jumbo CDs, institutional money funds and similar items.
Central banks are also cooperating partnerships, as well as being engines of inflation.
Lastly and back on the thread topic, what's your take on the most recent chart and conclusion? Have the deflationistas blown it with their Japan comparisons?
I had to chuckle a bit on the font sizes too, since my very first charts had a smaller default font... and I got flack from the 40 & over crowd due to vision issues. I just bring it up to illustrate that I'm damned if I do and damned if I don't sometimes.
Yeah, I hesitated a bit before writing that.
Already prefer your v6 graph above. Power to your shoulder dear Bart.
Have the deflationistas blown it with their Japan comparisons?
Totally! :)
Thanks for the charts bart. Great material.
Totally! :)
Thanks for the charts bart. Great material.
My pleasure Symbols.
I posted it on a few other forums today and imagine its another bee in the bonnet of folk like Mish by now.
babbittd
01-27-09, 06:54 PM
Hi Bart, another question for you. (I owe you 3 or 4 answers!)
I was thinking about this while driving and listening to political threater on the radio.
Can the 'stiumulus bill' be more aptly described as the 'inflation bill?"
The Left and the Right and the middle are arguing about the makeup of this bill. i.e. tax cuts, infrastructure spending, medicare, food stamps. Does it not matter what the money is spent on if the goal of the administration is simply, inflation?
Hi Bart, another question for you. (I owe you 3 or 4 answers!)
I was thinking about this while driving and listening to political theater on the radio.
Can the 'stimulus bill' be more aptly described as the 'inflation bill?"
The Left and the Right and the middle are arguing about the makeup of this bill. i.e. tax cuts, infrastructure spending, medicare, food stamps. Does it not matter what the money is spent on if the goal of the administration is simply, inflation?
Given your constraints on the goal being inflation only, yes it does matter but only in the sense that the spending needs to emphasize fast entry of the money into the economy.
It is better labeled as an inflation bill too... much like the original TARP was first called a bailout, and then the PR people started calling it a rescue bill.
babbittd
01-28-09, 01:12 AM
Given your constraints on the goal being inflation only, yes it does matter but only in the sense that the spending needs to emphasize fast entry of the money into the economy.
I wonder if the velocity of money is what Summers was referring to on Meet the Press when he talked about moving as quickly as possible to jolt the economy.
Glenn Black
02-02-09, 06:15 PM
Fourier series analysis is the standard technique for determining lead-lag correlations between two data sets.
For non-phased datasets, standard ANOVA (Analysis of Variance) can be used to determine the linear and non-linear correlation, and auto-correlation between data sets.
I wonder if the velocity of money is what Summers was referring to on Meet the Press when he talked about moving as quickly as possible to jolt the economy.
I doubt it since velocity isn't usually something that politicians talk about, but I don't know.
Fourier series analysis is the standard technique for determining lead-lag correlations between two data sets.
For non-phased datasets, standard ANOVA (Analysis of Variance) can be used to determine the linear and non-linear correlation, and auto-correlation between data sets.
True, and Excel has those capabilities available, but I'm not at all sure that the data quality and accuracy is high enough to bring out big guns like them... especially given that some folk associate tools like them with discredited Wall St. models.
vBulletin® v3.7.0, Copyright ©2000-2009, Jelsoft Enterprises Ltd.