Improving on the Permanent Portfolio

A place to talk about speculative investing ideas for the optional Variable Portfolio

Moderator: Global Moderator

User avatar
MachineGhost
Executive Member
Executive Member
Posts: 10054
Joined: Sat Nov 12, 2011 9:31 am

Improving on the Permanent Portfolio

Post by MachineGhost » Sun Nov 13, 2011 9:10 am

Intuitively, the HBPP concept made sense to me, but I've always had a nagging feeling in my gut that the risk of the portfolio was understated for purposes of "dumbed down" simplicity for conveying the concept to mainstream investors.  In no way shape or form could a simplistic allocation of 25%x4 be optimal in terms of balancing reward with risk.  The issue, then, becomes one of having enough historical data and at the proper granularity level to model the HBPP throughout various economic scenarios.  Going back to only 1972 and only annually on Simba's spreadsheet just doesn't cut it.  But needless to say, finding daily total return historical data for decades on end is nigh impossible (unless you have several thousand dollars to spend).  You'll have to forgive me if I didnt't take at face value HB's allegation that they tested a century of data on a supercomputer (where's the evidence?).  Plus most are aware of exactly how primitive the state of financial data and computing was in the late 70's compared to today.  Now on to the meat...

[img width=600]http://img690.imageshack.us/img690/368/ ... age001.gif[/img]

[img width=600]http://img502.imageshack.us/img502/368/ ... age001.gif[/img]
1939 to 1951 is hard to characterize.  It seems to me that "depreciation without stagnation" would be the best label.... obviously related to World War II and Bretton Woods.  The bond market did not respond to higher inflation in a positive or negative manner.  Growth was very robust in terms of stock gains.

[img width=600]http://img521.imageshack.us/img521/368/ ... age001.gif[/img]

[img width=600]http://img847.imageshack.us/img847/368/ ... age001.gif[/img]
Traditional 60% stocks, 40% bonds mix.

[img width=600]http://img36.imageshack.us/img36/368/pp ... age001.gif[/img]

[img width=600]http://img249.imageshack.us/img249/368/ ... age001.gif[/img]
The optimization eliminates T-Bills completely and ups the stock to about 43% and the bond to 31%.  It doesn't look it, but the depth and duration of drawdowns are lower than on all the other charts.

Notes: 1928-2010.  All portfolios are scaled to $100 in 1970 and were rebalanced annually.  No timing bands were used.  Taxes and
transaction costs were not included.

[img width=600]http://img197.imageshack.us/img197/9687/chartvh.png[/img]
Nominal returns for 25x4 since 1/1/1990 on a daily granularity basis, rebalanced annually.  CAGR was 6.82% and maximum drawdown was 15.05%.

Verdict: It seems probable that gold would have responded favorably during 1939 to 1951 if not controlled by the state, which is a strong testament to the fact that the so-called legality or illegality should never play a role in owning it.  Clearly, the lack of a free market in gold really hurts the HBPP on a real return basis.  82 years is still not enough history to provide enough instances where T-Bills and gold come into play to form a more optimal allocation.  At this point, it is far more productive to spend time on enhancing the bottom up than the top down.

MG
Last edited by MachineGhost on Tue Sep 16, 2014 10:03 am, edited 1 time in total.
"All generous minds have a horror of what are commonly called 'Facts'. They are the brute beasts of the intellectual domain." -- Thomas Hobbes

Disclaimer: I am not a broker, dealer, investment advisor, physician, theologian or prophet.  I should not be considered as legally permitted to render such advice!
Indices
Executive Member
Executive Member
Posts: 245
Joined: Sun Apr 25, 2010 10:51 pm
Contact:

Re: Improving on the Permanent Portfolio

Post by Indices » Sun Nov 13, 2011 9:44 am

It's all guessing when it comes to gold's theoretical performance during the 1930s-1960s. Remember: past performance does not guarantee future returns. The HBPP works for reasons other than past returns.
User avatar
MediumTex
Administrator
Administrator
Posts: 9096
Joined: Sun Apr 25, 2010 11:47 pm
Contact:

Re: Improving on the Permanent Portfolio

Post by MediumTex » Sun Nov 13, 2011 4:08 pm

The PP is built for a post-gold standard world.

In a gold standard world, I would expect another allocation method to work better than the PP.

It's also important to note that a 4x25% PP will actually spend very little time at 4x25%.  As soon as you set it up the various assets will typically immediately begin moving in different directions.
Q: “Do you have funny shaped balloons?”
A: “Not unless round is funny.”
D1984
Executive Member
Executive Member
Posts: 730
Joined: Tue Aug 16, 2011 7:23 pm

Re: Improving on the Permanent Portfolio

Post by D1984 » Sun Nov 13, 2011 6:04 pm

A couple of notes on performance of a hypothetical 4x25 PP in the late 1920s through early 1950s era that might shed some additional light on the poor "real" (inflation-adjusted) performance during the late 30s through 1951:

One, as mentioned earlier in this thread, gold prices were controlled by the government during this period. This may have helped during 1933 (when gold was "revalued" by FDR from $20 to $35 an ounce) but this help came when it was least needed (1933 was a bull year for stocks and there was still some deflation going on during this year IIRC so even if the PP had returned nothing it would still have had a positive real return). When it actually WAS needed (in the inflationary postwar years...roughly 1946-1948 in the US) gold stayed at $35 an ounce instead of increasing in value at several times the rate of inflation (which is what it would be expected to do during any period when inflation was in high-single or low double digits and interest rates were at sub-2% levels so one could get no real return from holding cash). However, if you look at silver prices in the immediate postwar era, they may or may not have increased enough (Clive, could you please repost your silver price data from 1945 to 1951) that using silver in lieu of gold here worked acceptably well. What I find interesting is that platinum prices increased (according to the USGS and Bureau of Minerals yearbooks) from $35 an ounce at the beginning of 1946 to $96 an ounce by the end of 1948. To be fair, 1927 through 1930 weren't exactly good years for platinum (neither was the 1937-38 recession...1928-1933 and 1937-1938 weren't too good for silver prices either; silver cratered from $0.58 to in 1928 to $0.25 in early 1933; recovered to roughly $0.58 by late 1936, and declined somewhat during 1937-1939 to the low $0.40 range) which is to be expected given that platinum and silver are both more "industrial use" precious metals than gold is. I'd like to see a PP chart for 1928-1968 (it could stop at 1968 because that's when the London Gold Pool $35 ceiling broke and we have a "free market" gold price to go by after that) that used a 50/50 mix of silver and platinum (so 12.5% of each total) in lieu of the 25% gold allocation.

Two, again as noted in this thread, interest rates were suppressed by the Fed from at least the beginning of WWII (1941 in the US) until the late 1940s. This means that t-bill yields (and interest rates in general short-term liquid cash equivalents like 1-3 yr treasuries) were pathetically low as compared to actual inflation during many years. Also, some states (Florida comes to mind) during this period levied an actual property tax (just like you pay if you own a house or land) on bank accounts, stocks, bills, bonds, etc but exempted Federal, state , and municipal securities. It wasn't much (less than 1/2 of 1% per annum) but it meant that someone might be willing to accept somewhat lower yields on t-bills rather than keep the money in a CD or savings account earning virtually nothing (CDs and passbook accounts at commercial banks during the later part of the Depression and through the late 1940s typically paid less than 1%) and then pay a property tax on it besides.

Three, note that most people (well, non-rich people who didn't have tens of thousands of $ to invest...which was a good bit of money even in the 1950s let alone the 1930s) back then didn't typically buy t-bills as a place to save or safely park some extra cash...there was obviously no such thing as TreasuryDirect or t-bill MMMFs during those years. They typically deposited their excess cash with a bank or building and loan (AKA a savings and loan or a building society). Interest rates paid on S&L accounts (which were FSLIC insured after 1934) exceeded t-bill rates considerably. For instance, rates paid on Federally-insured S&L shares (these are weighted annual averages) from 1935 to 1951 were as follows:

1935 – 3.69%

1936 – 3.44%

1937 – 3.48%

1938 – 3.49%

1939 – 3.39%

1940 – 3.25%

1941 – 3.13%

1942 – 3.02%

No date for 1943 or 1944...assume 1945 rates of 2.45%

1945 – 2.45%

1946 – 2.36%

1947 – 2.38%

1948 – 2.43%

1949 – 2.51%

1950 – 2.52%

1951 – 2.58%

For comparison, t- bill yields in 1935 were approximately 0.2%, briefly spiked to around 0.5% in 1937, and then fell to 0.1% or less by 1941. The Fed kept (by buying most newly issued t-bills) the rate at 0.375% during WWII and immediately after but it let rates rise starting in the late 1940s; by 1951 3-month t-bill rates had risen to around 1.55%.

Also, you might try substituting US Postal Savings deposits (paid 2% annually and were government insured) or Savings bond (paid 2.9% annually and tax-deferred...were for a 10-yr term but could be cashed in early with interest penalty but no loss of principal) for some of the t-bills...this wouldn't help a whole lot more in 1946-1948 but getting paid 2% or more during 10% inflation is still better than getting paid less than 1/2% during the same inflation.

Fourth, consider that LT treasuries might not have been as effective as they could have been during the Depression because:

A. 30-year bonds weren't available (most PP backtests use 10-year bonds for this period but the Federal Reserve has data for 12 to 16 year constant maturity yields for this period available so I'd use those instead)

B. LT treasuries didn't always perform totally as a "100% safe asset that everyone flees to in a depression/financial crisis" (ala 2008) since we were on the gold standard and that meant that we couldn't just print up money to pay our bonds and interest if needed (look at 1931-when the KreditAnstalt collapsed, gold was leaving the US, every country's banking system was in crisis, tax revenues were plummeting, and the US government didn't even know if it was going to be able to roll over the maturing WWI Liberty Bonds-as a good example of this happening)

C. Many US treasury securities were partly Federal tax-exempt until late 1941. This may have kept yields on LT treasuries from contributing as much as they could have (bond returns being a combination of yield plus capital appreciation or loss upon sale/rebalancing) been since rich investors who faced a 70% or more tax bracket (the top bracket in the late 1930s) might choose to buy US treasuries (and thus bid yields down more than would usually happen in a free market) rather than municipals. The Federal Reserve data bears this out; US LT bond yields were lower than muni yields from the mid 1920s though 1941 but the opposite was true from 1942 onwards. Lower yields due to this effect would help (lower bond yields=higher prices for current holders of bonds) bondholders during the 1930s (although I doubt it was enough to offset A and B above) but begin to hurt (by starting from a lower yield base than would be the case if Federal LT bonds had no tax advantages) them when yields began their long rise (lasting on-and-off until the early 1980s) after WWII.

Finally, Clive, I would like to know where you got timberland prices (either for the US or the UK) for the 1930s...the only timberland price or annual return data I have is NCREIF back to 1987 and Hancock Timber Index back to 1960.
Gumby
Executive Member
Executive Member
Posts: 4012
Joined: Mon May 10, 2010 8:54 am

Re: Improving on the Permanent Portfolio

Post by Gumby » Sun Nov 13, 2011 7:16 pm

I don't understand the point of this exercise. It wouldn't have made any sense to own a Permanent Portfolio before 1970. Every dollar in your pocket was already equal to a specific amount of gold. A Permanent Portfolio before 1970 would have been akin to holding 50% cash.

We don't need any data to tell us that a portfolio of 50% cash underperformed the market.
Nothing I say should be construed as advice or expertise. I am only sharing opinions which may or may not be applicable in any given case.
TripleB
Executive Member
Executive Member
Posts: 882
Joined: Sun Mar 27, 2011 1:28 am
Contact:

Re: Improving on the Permanent Portfolio

Post by TripleB » Mon Nov 14, 2011 8:16 am

One can very easily build a PP modification that will outperform the standard PP historically.

One cannot easily build a PP mod that will outperform going forward.
Exocet
Associate Member
Associate Member
Posts: 26
Joined: Sat Jun 05, 2010 1:51 am

Re: Improving on the Permanent Portfolio

Post by Exocet » Mon Nov 14, 2011 9:43 pm

There is no question that the PP has historically under-performed some allocations.  On the other hand, there is also no question that the PP's Sharpe ratio is by far the best of any traditional asset allocation techniques or traditional portfolios (with the exception of 100% cash allocation!).  Therefore, the PP offers the best risk-reward performance, which is not the same as saying "the best performance" - it is the best performance for the low level of risk assumed.  Let's keep in mind that HB (just like Buffett) would obsessively look at protecting principal.  That's why PRPFX is considered a "conservative" portfolio by Morningstar.

To me the PP reflects the first principle of Economics, one that is learned in the first lesson of Economics 101: we have to optimize, not maximize.  PP offers optimization of reward for the risk undertaken. 
Exocet
Associate Member
Associate Member
Posts: 26
Joined: Sat Jun 05, 2010 1:51 am

Re: Improving on the Permanent Portfolio

Post by Exocet » Mon Nov 14, 2011 9:48 pm

Clive wrote: If 35% stocks, 65% inflation bond type allocations can provide a more constant return over multiple time periods - then if you use a PP as the 'inflation bond' holding :-
It would be interesting to calculate the standard deviation of both allocations - that would give an indication if the PP's lower performance is the result of lower risk.  Let's not forget that over very long periods of time, any portfolio with high level of stock allocation will outperform any portfolio with low stock allocation (like PP's).  However, how many sleepless nights would any one can sustain with a portfolio of high stock allocation?
User avatar
MediumTex
Administrator
Administrator
Posts: 9096
Joined: Sun Apr 25, 2010 11:47 pm
Contact:

Re: Improving on the Permanent Portfolio

Post by MediumTex » Mon Nov 14, 2011 9:56 pm

Exocet wrote:
Clive wrote: If 35% stocks, 65% inflation bond type allocations can provide a more constant return over multiple time periods - then if you use a PP as the 'inflation bond' holding :-
It would be interesting to calculate the standard deviation of both allocations - that would give an indication if the PP's lower performance is the result of lower risk.  Let's not forget that over very long periods of time, any portfolio with high level of stock allocation will outperform any portfolio with low stock allocation (like PP's).  However, how many sleepless nights would any one can sustain with a portfolio of high stock allocation?
35/65 stocks/bonds is about what Vanguard Wellesley uses, and Wellesley has been around since 1970 so we have a simple proxy for this allocation to look at.

http://finance.yahoo.com/q/pm?s=VWINX+Performance

If you compare the PP to the performance in the link above, you will see that for much of the 1970s the PP easily beat this allocation both in nominal and real terms.
Q: “Do you have funny shaped balloons?”
A: “Not unless round is funny.”
User avatar
MachineGhost
Executive Member
Executive Member
Posts: 10054
Joined: Sat Nov 12, 2011 9:31 am

Re: Improving on the Permanent Portfolio

Post by MachineGhost » Wed Nov 16, 2011 11:04 pm

Fascinating contributions, guys!

I forgot to mention that in real terms, the maxium drawdown in the 25x4 during 1938 to 1951 was around 25%.  That's a jaw-dropping amount of net worth & purchasing power to have lost, especially outside the actual Great Depression.  Using the S&L rates for the given years reduces the real maximum drawdown to around 20%.  Still too high for a "conservative" portfolio.

So it seems even more clear to me now that "T-Bills" and "Gold" should be relabeled "Short Bonds" and "Real Assets" and highly benefit from active management.  Being a purist to a portfolio designed in hindsight for a post-gold standard economy seems non-robust if government intervention or human stupidity again prevents the market from achieving equlibrium in either of the two proscribed assets.

For "Short Bonds", I can think of CD ladders, grade A P2P lending, high quality corporate bonds, high quality municipal bonds (oxymoron?), I-Bonds, tax liens, Max Yield strategy.  They key would be keeping both the investment sizes small and maturities short so any losses have a minuscle impact.  A 10-year bond ladder has never lost money in any year in nominal terms, so the sweet spot in real returns for this category is likely to be between 2.5 and 5 years of duration.

For "Real Assets", I can think of semi-numismatic pre-1933 gold coins, pre-1965 junk silver coins, [colored] diamonds, rare stamps, rare books, rare autographs, improved land, farmland, "beach front" real estate, oil and gas wells, and, of course, timber.  The problem here can be the difficulty in rebalancing.

MG
"All generous minds have a horror of what are commonly called 'Facts'. They are the brute beasts of the intellectual domain." -- Thomas Hobbes

Disclaimer: I am not a broker, dealer, investment advisor, physician, theologian or prophet.  I should not be considered as legally permitted to render such advice!
User avatar
MediumTex
Administrator
Administrator
Posts: 9096
Joined: Sun Apr 25, 2010 11:47 pm
Contact:

Re: Improving on the Permanent Portfolio

Post by MediumTex » Wed Nov 16, 2011 11:41 pm

MachineGhost wrote: Fascinating contributions, guys!

I forgot to mention that in real terms, the maxium drawdown in the 25x4 during 1938 to 1951 was around 25%.  That's a jaw-dropping amount of net worth & purchasing power to have lost, especially outside the actual Great Depression.  Using the S&L rates for the given years reduces the real maximum drawdown to around 20%.  Still too high for a "conservative" portfolio.
As I noted above, the PP was not intended for use during periods prior to the 1970s that are being backtested above.  I wouldn't want anyone to get the wrong idea about the relevance of these backtesting results, so I want to make this point again.
So it seems even more clear to me now that "T-Bills" and "Gold" should be relabeled "Short Bonds" and "Real Assets" and highly benefit from active management.  Being a purist to a portfolio designed in hindsight for a post-gold standard economy seems non-robust if government intervention or human stupidity again prevents the market from achieving equilibrium in either of the two proscribed assets.
Part of living in a post-gold standard world is that we should take as a given a set of activist, misguided and frequently stupid economic, monetary and fiscal policies.  The PP is built with this state of affairs in mind, and has performed well through all manner of political stupidity in the last four decades.
For "Short Bonds", I can think of CD ladders, grade A P2P lending, high quality corporate bonds, high quality municipal bonds (oxymoron?), I-Bonds, tax liens, Max Yield strategy.  They key would be keeping both the investment sizes small and maturities short so any losses have a minuscle impact.  A 10-year bond ladder has never lost money in any year in nominal terms, so the sweet spot in real returns for this category is likely to be between 2.5 and 5 years of duration.
Anything but treasuries for the cash allocation is going to introduce counterparty and interest rate risk.  Some of the suggestions above are, IMHO, not suitable for "money you can't afford to lose."
For "Real Assets", I can think of semi-numismatic pre-1933 gold coins, pre-1965 junk silver coins, [colored] diamonds, rare stamps, rare books, rare autographs, improved land, farmland, "beach front" real estate, oil and gas wells, and, of course, timber.  The problem here can be the difficulty in rebalancing.
Take a look at how most of these assets did in 2008 compared to gold and then look at how these assets have done compared to gold over the course of the current 11 year bull market for gold.  IMHO, these assets are not suitable substitutes for gold in any way.  Even numismatic gold coins ask you to pay large premiums that are unnecessary when bullion will do the same job.

***

I appreciate your alternative views on the nature and role of a PP-like allocation, but as I have said many times the PP is what you might call "pre-optimized", and I have yet to see a reliable set of tweaks that will get you more safety or more reliable returns than the standard HB PP recipe. 

PRPFX is probably the most well-known set of HB PP tweaks, and for all its modifications it has done about the same as the HB PP over the course of its life, but with higher expenses and more volatility than the HB PP.  So far in 2011, the HB PP is beating PRPFX pretty decisively.

Once you have used the PP with your own money for a while and seen the solid returns and the consistently low stress levels that go with it, I believe the idea of filling your PP with municipal bonds, tax liens, rare autographs and beach front property will seem ghastly.

From a theoretical perspective, however, I understand that the urge to tinker is part of what makes us human, and it is an impulse that we can never completely outrun, even when we are dealing with an investment strategy that requires no tinkering to perform to specifications.
Last edited by MediumTex on Wed Nov 16, 2011 11:47 pm, edited 1 time in total.
Q: “Do you have funny shaped balloons?”
A: “Not unless round is funny.”
User avatar
Ad Orientem
Executive Member
Executive Member
Posts: 3483
Joined: Sun Aug 14, 2011 2:47 pm
Location: Florida USA
Contact:

Re: Improving on the Permanent Portfolio

Post by Ad Orientem » Thu Nov 17, 2011 1:12 am

I would like to add a quick point in support of MediumTex's preceding post.  Gold was chosen as the "hard asset"for the PP because it has for 6000 years served as a store of value and money.  Many of the other proposed "real assets" would fall more under the heading of "collectibles" than a universally recognized extra-national currency that is impervious to the depredations of the printing press.  Collectibles may make very good speculative investments in certain circumstances but their worth is likely to be heavily influenced by factors different than what drives gold.  In a severe economic crisis people may not be inclined to trade goods and services for rare stamps or works of art.  And real estate is neither conveniently portable nor terribly liquid.  Gold though (with apologies to Ben Bernanke) is money. 
Trumpism is not a philosophy or a movement. It's a cult.
User avatar
stone
Executive Member
Executive Member
Posts: 2627
Joined: Wed Apr 20, 2011 7:43 am
Contact:

Re: Improving on the Permanent Portfolio

Post by stone » Thu Nov 17, 2011 4:02 am

Clive, would daily review but a limit of one rebalancing in a given direction per 12 months also have given just a -25% rather than -50% draw down?
"Good judgment comes from experience. Experience comes from bad judgment." - Mulla Nasrudin
User avatar
stone
Executive Member
Executive Member
Posts: 2627
Joined: Wed Apr 20, 2011 7:43 am
Contact:

Re: Improving on the Permanent Portfolio

Post by stone » Thu Nov 17, 2011 7:24 am

Clive, I'm probably in a muddle but it sounds like you are projecting upon the PP some model of randomness that really doesn't fit reality. Basically supposing that the real values of LTT, gold, stocks and cash all fall together just as often as you would expect based on the product of the probablities of each of  them falling??
"Good judgment comes from experience. Experience comes from bad judgment." - Mulla Nasrudin
User avatar
stone
Executive Member
Executive Member
Posts: 2627
Joined: Wed Apr 20, 2011 7:43 am
Contact:

Re: Improving on the Permanent Portfolio

Post by stone » Thu Nov 17, 2011 9:36 am

Clive, am I understanding correctly that because you saw -28% in the annual data you compiled, you made the estimate that that implied that there may well have been a -50% loss intra-year that partially corrected to give a final -28% for the year?
"Good judgment comes from experience. Experience comes from bad judgment." - Mulla Nasrudin
User avatar
stone
Executive Member
Executive Member
Posts: 2627
Joined: Wed Apr 20, 2011 7:43 am
Contact:

Re: Improving on the Permanent Portfolio

Post by stone » Thu Nov 17, 2011 10:45 am

Clive, I guess a more direct translation of the Talmud might be 33% venture capital, 33% farmland, 33% reserves?? Back then, reserves meant silver but now it could be a PP?? A PP could in its entirety be thought of as reserves since it can be sold up at any time?
"Good judgment comes from experience. Experience comes from bad judgment." - Mulla Nasrudin
cabronjames

Re: Improving on the Permanent Portfolio

Post by cabronjames » Thu Nov 17, 2011 1:58 pm

@MachineGhost, I'm interested in your monthly PP return data set since 1990.

Is the monthly return data including dividend-reinvestment?

Do you have the component monthly return data for each asset: 30 yr T-Bond, cash, stock, gold?  Could you post a link to it

I want to find a source for div-reinvestment monthly return data for 30 yr T-Bond & cash that goes back as far as possible to 1973, even back to 1990 would be great.

I found a Fed Reserve source for monthly yield data, but thus far it is unclear how to approximate returns from this data set http://gyroscopicinvesting.com/forum/ht ... ic.php?t=9
gap
Full Member
Full Member
Posts: 60
Joined: Sat Nov 19, 2011 9:49 am

Re: Improving on the Permanent Portfolio

Post by gap » Sat Nov 19, 2011 10:48 am

Your results are remarkable. One comment and a request.

1. Do you or anyone else have monthly data on LT Bond prices  and Gold going back to the 20's or even 50's that you are willing to share

2. If interest rates go significantly higher my assumptions are
   Gold prices  go down
   LT Bond prices go down
   Stock prices  go down

The only asset that provides some increase is Cash (perhaps in the form of CDs or others).
So eliminating Cash may be undesirable

Are these reasonable assumptions and inferences?
Last edited by gap on Sat Nov 19, 2011 10:56 am, edited 1 time in total.
User avatar
MachineGhost
Executive Member
Executive Member
Posts: 10054
Joined: Sat Nov 12, 2011 9:31 am

Re: Improving on the Permanent Portfolio

Post by MachineGhost » Sat Nov 19, 2011 2:44 pm

Anything but treasuries for the cash allocation is going to introduce counterparty and interest rate risk.  Some of the suggestions above are, IMHO, not suitable for "money you can't afford to lose."
It seems to me counterparty risk is everywhere in the portfolio, short of holding physical possession of gold (and then you have other kinds of risk).  The volatility-capturing effect thrives on risk, so why shy away from it on an asset that is hard to have volatility to begin with?  I will argue that having continous cash flows and low durations by laddering would mitigate both interest rate, inflation risk and counterparty risk.  I need to be convinced that T-Bills alone justify the lost opportunity costs of higher income, especially in an extended deflationary scenario like Japan, and most especially from a government with outstanding total liabilities of 400% of GDP.  Would you argue that T-Bills are paying you a interest rate commensurate with that risk?  It seems to me the gold is the hedge against government, not the T-Bills.  Besides, gold has a limited window of opportunity to function as a medium of exchange only as long as the financial system infrastructure does not implode.  Otherwise, as the experience in the Vietnam War showed, soap and cigarettes will be more useful than a lump of shiny metal.  So long as we rely on the infrastructure not imploding, it seems silly to me to exclude assets that are as much a part of the infrastructure as T-Bills are.
Take a look at how most of these assets did in 2008 compared to gold and then look at how these assets have done compared to gold over the course of the current 11 year bull market for gold.  IMHO, these assets are not suitable substitutes for gold in any way.  Even numismatic gold coins ask you to pay large premiums that are unnecessary when bullion will do the same job.
One would have to consider the short-term risks of disinflation-type scenarios such as 1981, 1994 and 2008 against inflation expectations underperformance or government intervention such as 1938-1951 or the 70's which is why I wanted to check the long-term history of the PP performance.  I am less concerned about nominal losses -- which are a mirage -- than real losses.  Gold also tanked in 2008 before recovering in the 4th quarter.  Should the artificial distinctions of year end calculations determine the ultimate winners and losers or would it better be analyzed on a rolling window basis?

BTW, I didn't suggest numismatic coins, I suggested semi-numismatic.  There is a huge difference and currently semi-numismatic premiums are historically low over buillion spot, about 10% and sometimes even lower than fad bullion coins in some cases.  Obviously, it will require having a larger skill set to do active management smartly that to be a passive buy and hold PPer.  But that goes without saying.
I appreciate your alternative views on the nature and role of a PP-like allocation, but as I have said many times the PP is what you might call "pre-optimized", and I have yet to see a reliable set of tweaks that will get you more safety or more reliable returns than the standard HB PP recipe. 
Well, I do not consider that I'm "tweaking" the PP.  I am concerned about real underperformance historically and what to do about forestalling it in the future in what is seemingly an oversimplified portfolio allocation for the masses.  In my sore long-term experience, buy and hold of any type extracts a price of inefficacy in terms of lower reward for the risk in exchange for being lazy. 
PRPFX is probably the most well-known set of HB PP tweaks, and for all its modifications it has done about the same as the HB PP over the course of its life, but with higher expenses and more volatility than the HB PP.  So far in 2011, the HB PP is beating PRPFX pretty decisively.
I was lead to understand that the PRPFX was a prequel cursor to the PP, so isnt the PP actually the simplified tweak to the PRPFX?
Once you have used the PP with your own money for a while and seen the solid returns and the consistently low stress levels that go with it, I believe the idea of filling your PP with municipal bonds, tax liens, rare autographs and beach front property will seem ghastly.
I believe you may be right if you want to rely on the Federal governent's seemingly infinite ability to borrow and spend so that currency, CD's, bank accounts and Treasury bonds are all "risk free" soft defaults.  So between what you list here and what you didn't list earlier tells me those may be the appropriate "cash enhancements".
From a theoretical perspective, however, I understand that the urge to tinker is part of what makes us human, and it is an impulse that we can never completely outrun, even when we are dealing with an investment strategy that requires no tinkering to perform to specifications.
I am fearful about the risk of losing wealth in real terms for extended periods with the "simplified" PP.  Please do not rationalize this concern away because you think I can't resist the ability to tinker.  I rather not deal with the stress of what I'm doing, but the stress of plowing money into a passive allocation that suffered terribly in real terms from 1938-1951 is a huge concern.

MG
"All generous minds have a horror of what are commonly called 'Facts'. They are the brute beasts of the intellectual domain." -- Thomas Hobbes

Disclaimer: I am not a broker, dealer, investment advisor, physician, theologian or prophet.  I should not be considered as legally permitted to render such advice!
User avatar
MachineGhost
Executive Member
Executive Member
Posts: 10054
Joined: Sat Nov 12, 2011 9:31 am

Re: Improving on the Permanent Portfolio

Post by MachineGhost » Sat Nov 19, 2011 3:00 pm

Ad Orientem wrote: I would like to add a quick point in support of MediumTex's preceding post.  Gold was chosen as the "hard asset"for the PP because it has for 6000 years served as a store of value and money.  Many of the other proposed "real assets" would fall more under the heading of "collectibles" than a universally recognized extra-national currency that is impervious to the depredations of the printing press.  Collectibles may make very good speculative investments in certain circumstances but their worth is likely to be heavily influenced by factors different than what drives gold.  In a severe economic crisis people may not be inclined to trade goods and services for rare stamps or works of art.  And real estate is neither conveniently portable nor terribly liquid.  Gold though (with apologies to Ben Bernanke) is money. 
It is my understanding that gold was in the PP to be a store of value for when real interest rates are negative (i.e. unit of value depreciation), not to function as a medium of exchange in an post-Apocalyptic scenario.  Anything can be used as a medium of exchange (like soap or toilet paper), but not everything can be a store of value.  Your argument for being exclusionary to other hard assets primarily seems to rest more on gold as being a medium of exchange.  Historically, we've never had a real gold standard that gold bugs seem to think existed in some blissful nirvana except under Napoleon.  Over the centuries, many other things have been used as a medium of exchange than gold.  Even pre-1933, the gold basically just stayed in bank vaults while the actual medium of exchange was warehouse receipts or notes.

I can agree that gold is money when defined as a store of value, but so are diamonds and a few other assets that have historically saved people's proverbial bacon in the real world, not theoretically.  So in this case, there should be less consternation about enhancing the "Short Bonds" allocation to be more than just T-Bills if "Real Assets" remains limited to historically useful stores of value.  The PP seems very vulnerable to deflationary scenarios and interest rate underperformance.  Did T-Bills alone save Icelanders?

MG
"All generous minds have a horror of what are commonly called 'Facts'. They are the brute beasts of the intellectual domain." -- Thomas Hobbes

Disclaimer: I am not a broker, dealer, investment advisor, physician, theologian or prophet.  I should not be considered as legally permitted to render such advice!
User avatar
MachineGhost
Executive Member
Executive Member
Posts: 10054
Joined: Sat Nov 12, 2011 9:31 am

Re: Improving on the Permanent Portfolio

Post by MachineGhost » Sat Nov 19, 2011 3:12 pm

I used the Vanguard funds since 1990 because the dividends were reinvested as opposed to ETF's.  However, the data was daily, just the reporting was done in monthly for the annual totals.  Yahoo will have the historical data at daily.

You will not find 30-year T-Bond monthly data except at the FRED site and not earlier than 1977 (you will have to adjust it to be total return, simply calculate the capital gain from the previous period figure and add the current interest divided by 12, use "Percent", "Monthly", "End of Period").  I haven't seen Vanguard data go past 1990 on a daily basis.  Never seen monthly, but there is yearly at Vanguard's site.

To get the annual T-Bond returns back to 1928, I linked 10 years, then 20 years, than 30 years from FRED.

MG
cabronjames wrote: @MachineGhost, I'm interested in your monthly PP return data set since 1990.

Is the monthly return data including dividend-reinvestment?

Do you have the component monthly return data for each asset: 30 yr T-Bond, cash, stock, gold?  Could you post a link to it

I want to find a source for div-reinvestment monthly return data for 30 yr T-Bond & cash that goes back as far as possible to 1973, even back to 1990 would be great.

I found a Fed Reserve source for monthly yield data, but thus far it is unclear how to approximate returns from this data set http://gyroscopicinvesting.com/forum/ht ... ic.php?t=9
"All generous minds have a horror of what are commonly called 'Facts'. They are the brute beasts of the intellectual domain." -- Thomas Hobbes

Disclaimer: I am not a broker, dealer, investment advisor, physician, theologian or prophet.  I should not be considered as legally permitted to render such advice!
User avatar
MachineGhost
Executive Member
Executive Member
Posts: 10054
Joined: Sat Nov 12, 2011 9:31 am

Re: Improving on the Permanent Portfolio

Post by MachineGhost » Sat Nov 19, 2011 3:18 pm

2. If interest rates go significantly higher my assumptions are
   Gold prices  go down
   LT Bond prices go down
   Stock prices  go down

The only asset that provides some increase is Cash (perhaps in the form of CDs or others).
So eliminating Cash may be undesirable
I agree that it is undesirable to eliminate the "cash".  But the "cash" doesn't increase in nominal terms because it has a 0 duration and it doesn't go up very much in real terms because it lacks leverage.  But do note that "cash" in the PP is not actaully cash, it is a fixed income/loan/Short Bond allocation.  Real cash is currency that would be stuffed under the proverbial mattresses and rot away over time from inflation without exception.  See blow.

Government backstopping currency, CD's, bank accounts, and T-Bonds doesn't change the nature of each of the assets.  One is a medium of exchange, two are loans to the private sector, the last is a loan to government.

MG

Image
Last edited by MachineGhost on Sat Nov 19, 2011 3:30 pm, edited 1 time in total.
"All generous minds have a horror of what are commonly called 'Facts'. They are the brute beasts of the intellectual domain." -- Thomas Hobbes

Disclaimer: I am not a broker, dealer, investment advisor, physician, theologian or prophet.  I should not be considered as legally permitted to render such advice!
D1984
Executive Member
Executive Member
Posts: 730
Joined: Tue Aug 16, 2011 7:23 pm

Re: Improving on the Permanent Portfolio

Post by D1984 » Sun Nov 20, 2011 2:33 am

To get the annual T-Bond returns back to 1928, I linked 10 years, then 20 years, than 30 years from FRED.
Ten-year T-bonds aren't really truly "long-term" bonds; look at what happened in the 1990s and 2000s when they were used in the Japanese PP (they worked as expected-prices rose as yields fell-but didn't show nearly enough punch in a deflationary scenario; certainly not as much as 20 or 30 year bonds would) or in 2008 in the American PP (owning TLT or owning 25-30 year bonds directly via TreasuryDirect provided you with an asset that gained enough to offset the stock losses; a ten-year bond almost certainly wouldn't have).

If you want slightly longer term T-bond data, it is available as follows:

For 1926-1930 - Table 128 and Table 129 of the 1919-1941 version of "Section 12: Money Rates and Securities Markets" at http://fraser.stlouisfed.org/publicatio ... tion12.pdf (warning....this is a huge PDF file) gives monthly yield data for all US government bonds maturing or callable in 12 years or more (note that that means these are bonds that have a MINIMUM of 12 years to maturity or call).

For January 1931-late 1938 - the December 1938 Federal Reserve Bulletin (pgs 1045-1046) has WEEKLY (even better and more specific than monthly) yields for US government bonds due or callable after 12 years. This document is located at http://fraser.stlouisfed.org/publicatio ... 121938.pdf . You actually only will be using these yields until the end of 1935 (see the next paragraph below for the reason why...basically even longer term yield data than 12 years is available after December 1935....but they go all the way to 1938)

For January 1936 to mid-1945 - the May 1945 Federal Reserve Bulletin (pgs 486-490) has weekly yields for US government bonds due or callable after a minimum of 15 years. This document is located at http://fraser.stlouisfed.org/publicatio ... 051945.pdf

For 1945-1953 - The 1941-1970 version of "Banking an Monetary Statistics: Section 12: Money Rates and Securities Markets" at http://fraser.stlouisfed.org/publicatio ... tion12.pdf has weekly average yield data (in table 12.12) for US government bonds with a minimum of 15 years until maturity or call; this data is from January 1941-March 1952. From April 1952-March 1953 it has weekly average yield data for all US government bonds with a maturity of 12 years or more (said average yield almost certainly includes some of the above bonds as well since a bond with 14 or 15 years to maturity by default has more than 12 years until maturity since 15 is more than 12).

if you just want 20-year yields (nothing with 12 yrs or 15 yrs left to maturity...just 20 years) monthly for all of 1953, they are available in Table 3.A.3; Column R20; of "Corporate Capital Structures in the United States; Debt and Equity Yields 1926-1980" by the NBER. It can be downloaded at http://www.nber.org/chapters/c11419.pdf

I also have available as an Excel file (if you want it) weekly yield data for May 1953-January 1955 for the 30-yr (issued at 3.25% yield; due in 1983; callable in 1978) issued on May 1, 1953. This data was taken from various Federal Reserve Bulletins from the mid-1950s (from February 1955-December 1957 the monthly Federal Reserve Bulletins provided a weekly blended average yield of the abovementioned 3.25% bond and a US government bond originally issed at 3.00% in February 1955 that was due in 1995...that's right folks, a marketable 40-yr US Government bond did exist at one point).

Finally, since (IIRC) either you or CabronJames mentioned needing monthly gold prices/returns before 1973, check out the LBMA's website at http://www.lbma.org.uk/pages/index.cfm? ... ld_fixings for DAILY historical free-market gold fix prices from early 1968 (when the Gold Pool broke and was unable to control the price at $35 any longer...this was the beginning of a free-market price in gold even though US citizens still technically couldn't legally own it) to the present. Prices are given in USD, pounds sterling, and (for the last decade or so) Euros.
User avatar
MediumTex
Administrator
Administrator
Posts: 9096
Joined: Sun Apr 25, 2010 11:47 pm
Contact:

Re: Improving on the Permanent Portfolio

Post by MediumTex » Mon Nov 21, 2011 12:02 am

I will just take a crack at a few of these:
MachineGhost wrote: I need to be convinced that T-Bills alone justify the lost opportunity costs of higher income, especially in an extended deflationary scenario like Japan, and most especially from a government with outstanding total liabilities of 400% of GDP.  Would you argue that T-Bills are paying you a interest rate commensurate with that risk?
In a post-gold standard world, I want to own debt of the entity with the power to print the money.  That entity can never default, except indirectly through inflation.
It seems to me the gold is the hedge against government, not the T-Bills.  Besides, gold has a limited window of opportunity to function as a medium of exchange only as long as the financial system infrastructure does not implode.
Gold provides you with the type of exposure you need under certain economic conditions.  That's all it's for.  Any post-TEOTWAWKI uses are just a bonus.
One would have to consider the short-term risks of disinflation-type scenarios such as 1981, 1994 and 2008 against inflation expectations underperformance or government intervention such as 1938-1951 or the 70's which is why I wanted to check the long-term history of the PP performance.
I still don't understand why you are backtesting during periods prior to 1971.  The PP is only designed to work in a post-gold standard world.  Testing it in a gold standard era is like testing a BMW on a dirt track--it's simply not designed for those conditions and shouldn't be expected to perform well in that environment.
BTW, I didn't suggest numismatic coins, I suggested semi-numismatic.  There is a huge difference and currently semi-numismatic premiums are historically low over buillion spot, about 10% and sometimes even lower than fad bullion coins in some cases.  Obviously, it will require having a larger skill set to do active management smartly that to be a passive buy and hold PPer.  But that goes without saying.
But what is the rationale for semi-numismatic coins in the first place?  Bullion gives you the exposure you need.  There is no need to assume the added risk of fluctuating premiums to spot price that semi-numismatic coins offer.
Well, I do not consider that I'm "tweaking" the PP.  I am concerned about real underperformance historically and what to do about forestalling it in the future in what is seemingly an oversimplified portfolio allocation for the masses.  In my sore long-term experience, buy and hold of any type extracts a price of inefficacy in terms of lower reward for the risk in exchange for being lazy.
Part of what is impressive about the PP, though, is that during the period for which it was designed--i.e., post-1971--it has performed well when virtually every other allocation has let investors down at some point.
I was lead to understand that the PRPFX was a prequel cursor to the PP, so isnt the PP actually the simplified tweak to the PRPFX?
Part of the essence of the PP is an agnostic view of the future.  PRPFX clearly tilts toward an inflationary view of the future, and this tilt has not provided a meaningful advantage over the HB PP since its inception.  In Harry Browne's writings, he makes it clear that PRPFX is a good fund (and is obviously one of the allocation methods he was writing about in the late 1970s and early 1980s, along with Terry Coxon), but the 25%x4 allocation is what he came to view as being safer and more consistent with the premise that the future is fundamentally unpredictable.  I guess you might say that there was a period of early experimentation in Harry Browne's thinking during which the PRPFX strategy and the 4x25% strategy were both offered as safe and stable approaches to investing.  As time passed, however, I think that Harry Browne came to see the 4x25% approach as being just as effective as the PRPFX approach, while offering more safety and stability under certain market conditions.
I am fearful about the risk of losing wealth in real terms for extended periods with the "simplified" PP.  Please do not rationalize this concern away because you think I can't resist the ability to tinker.  I rather not deal with the stress of what I'm doing, but the stress of plowing money into a passive allocation that suffered terribly in real terms from 1938-1951 is a huge concern.
The PP wold not have been appropriate for "money you can't afford to lose" in the 1938-1951 time frame.  Since 1971, however, the PP has proven to be a near perfect location for "money you can't afford to lose." 

***

As I have said many times before, the PP is not for everyone, and it may not be for you.  As you are working through this matter, though, I would try not to be too distracted by what a theoretical PP would have done during any period prior to 1971, for the same reasons that you wouldn't want to rule out buying a BMW because you heard that it didn't perform well off-road.

I would instead look at the PP's performance in both real and nominal terms since 1971, and compare is to other allocation methods over that period.  If you decide not to use the PP going forward, an interesting experiment is to track your own portfolio against a hypothetical PP (along with the volatility of the two portfolios), just to see how the two compare.  It was this exercise that initially got my attention when I first became interested in the PP.

The PP is certainly not the only game in town--it's just the one that makes the most sense to me.  Once you get used to sleeping well at night, it's hard to go back.  If the PP ever does stop working, though, I will be the first one to say let's all look for something better, but until that happens I am just going to enjoy the ride.

One approach for even the most hardcore skeptic is to allocate a small part of your portfolio to the PP for six months or so, and just see how it goes.  See how you like the way it operates in real-time and with real money.  That will often tell you more than any amount of backtesting or theoretical analysis ever could.
Q: “Do you have funny shaped balloons?”
A: “Not unless round is funny.”
magneto
Full Member
Full Member
Posts: 56
Joined: Fri Apr 08, 2011 8:30 am

Re: Improving on the Permanent Portfolio

Post by magneto » Mon Nov 21, 2011 9:49 am

Clive wrote :-

Ideally you want to be holding assets that have high volatility, inverse correlation and individually provide positive real (after inflation) gains through price appreciation and/or income.

unquote


The mention of inverse correlation triggered a thought that has been confusing me for some time.  With two asset classes A and B, do you really want B to decline by the same amount when A is appreciating?

Would not a correlation of 0 be better than -1.00, such that B retains its value be more of a benefit to the portfolio.

Someone tell me why my thinking is off.
Post Reply