Correction Ahead

[MachineGhost]

As some here know, I've shorted three of the last three corrections successfully, but due to poor choice in financially engineered vehicle selection, I did not make out like a bandit as expected.

Once again we are on the cusp of another correction.  Whether it occurs here in the seasonally bearish period or later in the year after suckering everyone in one last time -- I do not know.  What I do know is this year is the top of several cycles and will usher in several years of worldwide deflation, probably due to China deleveraging.  China is more leveraged now than the US was in 2007 or Japan in 1989.  Take heed.  This is your final warning.

What I also know is that I'll be smart this fourth time and just buy long-term bonds.  I don't have faith the PP will weather a simultaneous assault on overvalued stocks, no demand gold and zero interest cash...  the dreaded Achilles Heel scenario.  While historically the world's core economy never hyperinflates, I think we're in a new era of unhinged credit recklessness and there is always a remote possibility the world will lose faith in even the U.S. Treasury during the crisis.  Already, the EU is resorting to negative nominal interest rates in a vain and futile attempt to to stimulate their economies. 

How much further depraved can all of this financial absurdity go?  We'll soon find out.  My technical indicators are still currently bullish, but the storm's a-comin'.

[modeljc]

MG, welcome back to the forum! 

Are you going to go 100% into long treasuries?  If so, when?

I would be interested to know your current allocation.  It sounds like you may not be  a PP believer.

[Peak2Trough]

Sounds like someone's been reading too much Marc Faber. 

[Stewardship]

Do you also do palm reading?

[murphy_p_t]

"As some here know, I've shorted three of the last three corrections successfully, but due to poor choice in financially engineered vehicle selection, I did not make out like a bandit as expected."

MG...why do you expect your selection to be better this time?

As an aside, it sounds like you've assigned 0% chance of collapse of petrodollar system and 100% chance that US$ remains the world's reserve currency well into the future?

[MachineGhost]

Are you going to go 100% into long treasuries?  If so, when?

I wouldn't go 100% long on Treasuries, but I would certainly put 25% into it, just without the stocks and gold draggin' down da returns.   A few percent here or there is quite easily to lose with the fully-invested PP.

I actually did buy shallow Treasuries last April or May on a buy signal that later turned out to be false due to noise in the data.  Guess when I sold?  Right at the bottom on Dec. 31 to impulsively take the tax loss.  So I'm naturally reluctant to go through that stress again.

On the other hand, the false buy made me take a closer look at my bond data and I found some discrepanciees that needed to be fixed and as a bonus, it is now pure 30-year CMT all the way back to 1941.  No more hacks.

[MachineGhost]

I would be interested to know your current allocation.  It sounds like you may not be  a PP believer.

I'm currently 61%-87.5% in cash depending on the portfolio in question.  No stock exposure in any portfolio other than a small percentage to gold stocks.  I'm not a "buy and hold" PP believer; but it is the strategic asset allocation that I use.  There may be slightly better ways to allocate the PP than 25x4 but the edge is overwhelmed by individual asset class risk.

[MachineGhost]

MG...why do you expect your selection to be better this time?

As an aside, it sounds like you've assigned 0% chance of collapse of petrodollar system and 100% chance that US$ remains the world's reserve currency well into the future?

Isn't that like asking why the PP works?

.5% chance in this current cycle if only because the rest of the world is more screwed up than "Rome".  Beyond that in the next cycle, I think we're looking at it as a factor.  Lets worry about that bridge when we come to it.

[annieB]

MG:

Thanks for your update.
We've missed you and your posts...

[Reub]

MG, DecisionMoose just switched their investment recommendation to long term bonds (EDV). They must be listening to you!

[MachineGhost]

MG, DecisionMoose just switched their investment recommendation to long term bonds (EDV). They must be listening to you!

Interesting!

The average overall gain for the Moose bond signals is 3.21% and the median is 2.13%.  Only .15% and -6.23% in 2008 crisis, but 5.12% in the 2010 correction, -.87% in the 2011 correction and 5.54% in the 2012 correction.  The 2010 signal came in the middle of the correction if I recall correctly, the 2011 signal came near the end, but the 2012 signal was at the beginning.  Since the Fed showdogged three times to stop the corrections, a signal near the beginning like this year is turning out seems the safest.

[dualstow]

MG, do you think gold is going to go down with stocks or just tread water?

[Reub]

I'd love to see a blow-off top in the equity markets soon before it crashes.

[dualstow]

I'd love to see a blow-off top in the equity markets soon before it crashes.

Yeah, I would have liked a little more bubble activity.

[MachineGhost]

MG, do you think gold is going to go down with stocks or just tread water?

Collapse.

But keep in mind the timing is different.  Stocks, gold, bonds all rally and collapse at different times during a correction.  Stocks collapse, gold goes up a bit then collapses, bonds rally and thats the end of it until there is an economic recovery or a 1930's-style sovereign bond implosion at which time they collapse.  Cash is king throughout.  Quantitative easing can keep the bond party going longer.

I think the real trick is recognizing when we've crossed the Nile into sovereign debt implosions.  And if that is still possible with quantitative easing.  I tend to lean more towards the Japanese experience.

[D1984]

Are you going to go 100% into long treasuries?  If so, when?

I wouldn't go 100% long on Treasuries, but I would certainly put 25% into it, just without the stocks and gold draggin' down da returns.   A few percent here or there is quite easily to lose with the fully-invested PP.

I actually did buy shallow Treasuries last April or May on a buy signal that later turned out to be false due to noise in the data.  Guess when I sold?  Right at the bottom on Dec. 31 to impulsively take the tax loss.  So I'm naturally reluctant to go through that stress again.

On the other hand, the false buy made me take a closer look at my bond data and I found some discrepancies that needed to be fixed and as a bonus, it is now pure 30-year CMT all the way back to 1941.  No more hacks.

Where did you obtain 30-year CMT Treasury data going back to 1941? CRSP daily data only goes back to 1961 or 1962 so far as I know (their monthly data goes back to 1941, though, IIRC).

Was this data daily, month-end, middle-of month, or just a monthly average? CRSP from 1941 to the early 60s was end-of month (either the 28th, 29th, 30th, or 31st) and Salomon Brothers has middle-of-month (data taken each month on the 15th) going back to 1945 but no one that I know of has daily 30-year CMT data going back to 1941.

Besides, there weren't actually any marketable 30-year bonds from around 1942 or 1943 to early 1953 (there was an issue in 1942 of a T-bond due in 1972 but optionally callable in 1967 and then there was a 30--or perhaps 40 year--but IIRC it was 30 years and non-callable) year bond issued in the spring of '53, but between those....nada...there were shorter-term bonds issued but during the intervening years but nothing as long as 25 or 30 years (the 1942 bond was reopened in 1945 but that was the same issue just re-opened for subscription). There was a non-marketable 29-year bond issued around the time of the Fed-Treasury "Treaty of Omaha" accord in early 1951 but that wouldn't be in the CMT data as it was non-marketable and non-traded.

I would be really interested in knowing where I can download/buy daily 30-year CMT going back to 1941.

[murphy_p_t]

MG, do you think gold is going to go down with stocks or just tread water?

Collapse.

But keep in mind the timing is different.  Stocks, gold, bonds all rally and collapse at different times during a correction.  Stocks collapse, gold goes up a bit then collapses, bonds rally and thats the end of it until there is an economic recovery or a 1930's-style sovereign bond implosion at which time they collapse.  Cash is king throughout.  Quantitative easing can keep the bond party going longer.

I think the real trick is recognizing when we've crossed the Nile into sovereign debt implosions.  And if that is still possible with quantitative easing.  I tend to lean more towards the Japanese experience.

Interesting scenario you paint...I haven't read any commentary which is bullish on sovereign debt. Any blogs or essays you can recommend that expand on your outlook? I've read too much gold-buggery doomer stuff, so I think I need to balance it out w/ some other outlook...

[MachineGhost]

Further regarding DecisionMoose, the maximum drawdown for all the bond trades is -21.91%.  It can be enhanced by the use of a 9.72% trailing stop that reduces it to -14.54% while improving the gain by 22%.

Overall, it's not a very impressive track record for buying a risky, zero coupon bond fund, but the best gain was from May 2000 to January 2001 at 18.09% (and .58% later that year, 19.45% in 2002 to 2003 and .76% in 2003).  None of the trades declined below its entry by more than about -5% except in 2011 when it reached almost -10%.

Overall, a 80% win rate and the three losing years were 2005, 2008, 2011.  Did the Fed do anything to juice the market in 2005 or 2008?

It pretty much missed the boat on the subprime bubble collapse with a net negative bond return after all is said and done.  The entire portfolio has sucked since that time as well.

[MachineGhost]

I would be really interested in knowing where I can download/buy daily 30-year CMT going back to 1941.

Dream on.  I had to generate the daily returns myself from whatever data was available, probably the weekly or monthly series from FRED.

[MachineGhost]

Interesting scenario you paint...I haven't read any commentary which is bullish on sovereign debt. Any blogs or essays you can recommend that expand on your outlook? I've read too much gold-buggery doomer stuff, so I think I need to balance it out w/ some other outlook...

I don't think you'll find anything since 99.99% are ideologues not objective realists about history and human behavior.  So I have nothing to suggest, but anyone with a PP knows bonds go up in a deflation, up to a certain point.  Aren't the doom porners aware of that or are they all betting on armegeddeon?

[D1984]

I would be really interested in knowing where I can download/buy daily 30-year CMT going back to 1941.

Dream on.  I had to generate the daily returns myself from whatever data was available, probably the weekly or monthly series from FRED.

Hmmmm.....so far as I can tell FRED (and the Federal Reserve) don't have continuous daily (or weekly) yield data going back to 1941 for any maturity as long as 20 or 30 years (and neither do the monthly FRBs). They have actual 20 year daily data from 1962 onwards, 20 year monthly average yield data (synthetically derived) from 1953 onwards, and 30-year daily yield data from 1977 onwards. There is a weighted average of the 30 year of 1953 and then later adding on the 40 year of 1955, from early 1953 to the end of 1957. Any other "long-term" yields (the Fed has these records weekly going all the way back to before the Depression) from 1947/1948 to early 1953 will be at most for maturities of 15 to 20 years or in some cases just over 13 or 14 years (by the 1951-52 period) because no longer-term marketable bond was outstanding by then...oh, and from January 1958 to at least 1962 the "long-term" average yield that was published weekly/monthly in the FRBs was an average of any bond over 10 years' maturity (as per the Fed's own words: "This series, representing yields on bonds due or callable in 10 years or more, replaces those shown in earlier BULLETINS"...this was from pg 84 of the January 1958 FRB) so it was hardly 30-year data despite the fact that bonds with a lot closer to 30 years' maturity were available then.

Also, how do you accurately derive daily CMT returns from monthly or weekly data? The actual real yield would vary day by day during the week/month.

If you want the honest 100% accurate daily yields I'm still saying the only way to obtain those is from CRSP for post-1961 data and from the WSJ bond pages (they list the market bond prices but yields can easily be computed from those given a known par value) day by day at a time on microfiche from the 1920s to 1961 (from 1900 to 1920 the Magazine of Wall Street published week-ending yields on government bonds each Friday if you want to go back even farther). Did you find your CMT data (daily derived from weekly/monthly yields) was a close enough approximation to work as intended despite being taken from less granular than daily data?

[MachineGhost]

Hmmmm.....so far as I can tell FRED (and the Federal Reserve) don't have continuous daily (or weekly) yield data going back to 1941 for any maturity as long as 20 or 30 years (and neither do the monthly FRBs). They have actual 20 year daily data from 1962 onwards, 20 year monthly average yield data (synthetically derived) from 1953 onwards, and 30-year daily yield data from 1977 onwards. There is a weighted average of the 30 year of 1953 and then later adding on the 40 year of 1955, from early 1953 to the end of 1957. Any other "long-term" yields (the Fed has these records weekly going all the way back to before the Depression) from 1947/1948 to early 1953 will be at most for maturities of 15 to 20 years or in some cases just over 13 or 14 years (by the 1951-52 period) because no longer-term marketable bond was outstanding by then...oh, and from January 1958 to at least 1962 the "long-term" average yield that was published weekly/monthly in the FRBs was an average of any bond over 10 years' maturity (as per the Fed's own words: "This series, representing yields on bonds due or callable in 10 years or more, replaces those shown in earlier BULLETINS"...this was from pg 84 of the January 1958 FRB) so it was hardly 30-year data despite the fact that bonds with a lot closer to 30 years' maturity were available then.

Also, how do you accurately derive daily CMT returns from monthly or weekly data? The actual real yield would vary day by day during the week/month.

If you want the honest 100% accurate daily yields I'm still saying the only way to obtain those is from CRSP for post-1961 data and from the WSJ bond pages (they list the market bond prices but yields can easily be computed from those given a known par value) day by day at a time on microfiche from the 1920s to 1961 (from 1900 to 1920 the Magazine of Wall Street published week-ending yields on government bonds each Friday if you want to go back even farther). Did you find your CMT data (daily derived from weekly/monthly yields) was a close enough approximation to work as intended despite being taken from less granular than daily data?

You're a pesky bugger!  I just checked my code and what I did was I used the proprietary daily T-Bond yield that starts in 1941 (likely from the WSJ) and then switched over to the FRED 30-year in 1977 and back to the WSJ during the 2002 to 2006 hiatus and then back to the FRED 30-year thereafter.  Unlike the last time we discussed this, there's no longer a 20-year substitution that causes a problem due to the yield differential when switching.  I refactor it every day to a 30-year CMT.

The end result is it turns out to be a waste of time trying to time the bond market technically.  The 1941-1980 bear was just so brutal, nothing could offset it.  Probably the relative strength approach that DM uses it the only way to eek some value out of bonds; otherwise stick to using bonds as a hedge for your equity.  With that becoming more important right now, I still think duration matching or whatnot is the ideal hedging ratio.  I never followed up on that.

Clearly, the PP works like layers in a pyramid and the first layer is stocks, then bonds, then gold and then cash at the apex.  If cash could be duration increased somehow, it would be the perfect offset to gold and then we could all sleep at night without worrying about that 25% maximum drawdown.

[dualstow]

Oh dear lord, gold seems to have dropped 2% this morning.
Thanks a lot, MG. 

EDIT: P.S.:

China Is Losing Its Taste for Gold {WSJ}
Economic Slowdown, Constrained Credit Markets Suppress Demand
http://online.wsj.com/news/articles/SB1 ... 0211955954

[kka]

It will be interesting to watch things play out this year.  On one hand, MachineGhost tells us bonds will rally while stocks and gold tank, and kshartle is convinced the exact opposite will occur (see http://gyroscopicinvesting.com/forum/bo ... /#msg92357).  I'll acknowledge I don't know what will happen and continue to hold all 3 as well as cash, rebalancing as needed.  So far this year, gold +8%, bonds +9%, stocks -1%.

[Kshartle]

It will be interesting to watch things play out this year.  On one hand, MachineGhost tells us bonds will rally while stocks and gold tank, and kshartle is convinced the exact opposite will occur (see http://gyroscopicinvesting.com/forum/bo ... /#msg92357).  I'll acknowledge I don't know what will happen and continue to hold all 3 as well as cash, rebalancing as needed.  So far this year, gold +8%, bonds +9%, stocks -1%.

* I consider anything under 1.5 years short term and short term guesses.......guesses.

I shun bonds and embrace gold and stocks based on what I expect over the next ten years.....not ten months.

We could very well have another 2008 (although I doubt the Fed will allow it). The point is even if we do I expect a bazooka of inflation to incinerate the bonds and propel Gold farther than almost anyone can expect.

The Central banks of the world are united in an inflation program. The financial media and hack economists repeat daily how wonderful inflation is for us to condition us for it. The Fed can print as much as they want. They can ensure a growing money supply all on their own.