Golden Butterfly Portfolio

[Tyler]

Good info, Desert.

I start in '75 since that was the first year gold was legal to own in the U.S.

I just wanted to point out that this is only partially true. 1975 was the first year that Americans could own or trade physical gold, but gold certificates were legal for individual investors since 1964. People just couldn't personally redeem those certificates for the physical product. Think of it as a simple version of modern gold ETFs.

https://en.wikipedia.org/wiki/Gold_Reserve_Act

UPDATE: It turns out that I was wrong on this point. I later learned that the legalization of gold certificates was about them as collectibles rather than any relationship with physical gold. So Desert is correct that gold was illegal to hold in the US from 1934-1975.

[barrett]

Barrett, if I remember correctly, you and I are at somewhat similar points in life, though your daughter is much closer to college than my son. But you and I might end up retiring around the same time (hopefully soon!). I've been studying up on the best ways to cash flow retirement, pre-SS. Tyler probably knows far more about this than I do, since he's already doing it and much earlier than me. Due to some odd past opportunities to stash money in pre-tax accounts, we have an unusually high percentage of our savings in traditional IRA's. So my focus on cash flow is getting funds out of those accounts with zero penalties and very low income taxes. It's going to be interesting. Anyway, back to your first point regarding gold ETF's in the IRA ... I agree that some gold in the IRA should reduce volatility/drawdowns of those accounts. I still always try to look at my entire portfolio as one, so I personally don't care that much about the volatility of any single account. Does that make sense?

Regarding gold's performance in the 2000's, I need to geek out on Tyler's article on correlation and then reply later.

Desert, You remember correctly. I will be 59 in September (still have the body of a 57-year-old!) and our daughter just finished her first year of college. Semi-retirement/part time work is already here for me through a combination of factors. And, yes, it does make sense what you are saying about looking at your portfolio as a whole. In our situation, my wife and I will hit retirement with a fairly even split between tIRAs and taxable (but with a very high cost basis on the taxable). Adding to the confusion is that we both have Roth accounts which we intend to let grow until age 70.5 (early 2029 for me and 2037 for her). Plus we have a big (for us) stash of savings bonds that are keepers until maturity. Maybe this is TMI but I just lay it all out because, for us, almost as important as how much we have saved is what order we draw from the accounts. My plan is to draw from taxable and tIRAs first to avoid the worst of RMD consequences and then, starting in 2029, withdraw from my Roth while starting to collect SS.

That's all a long way of saying that I'd like for our tIRAs and taxable accounts to perform more or less the same so that I don't have to be concerned about "selling low" in either of them. But maybe I am just over thinking things.

And I get what you are saying about gold's price from 1972 through 1980. We can never know for sure how much of the run up was due to depegging, how much was due to speculation and how much was due to financial uncertainty/inflation. To Tyler's credit, his site has the start-date sensitivity info so we can discount the 1970s if we want to. But the run up during the 2000s can't be ignored. In general, people who seem to have strong feelings about the 1970s, seem to be less sure about how to frame gold's performance during the 2000s. I think understanding that period beyond Tyler's "lack of correlation to stocks" is important for those of us who want to hold a chunk of gold. Maybe it's all been hashed out on here in other threads before I arrived in early 2014. I just haven't seen a satisfying discussion of it.

I think I got a little over-caffeinated this morning. Sorry about that!

[sophie]

My main struggle is whether to make the pp a larger part of my total, which effectively means a bet on gold, buying more gold. What if I load up the truck and gold doesn't come back in my lifetime? What if I don't load up, and merely break even when gold soars once again?

This is from the other gold thread (the one where I was hoping kick off a discussion of the wisdom of holding a ton of gold in ETFs), but it better belongs here I think.

It's a really good question. You could of course ask the exact same question about any other asset. Stocks can be down or flat for periods up to 20 years, which can qualify as "lifetime" for some of us. And let's not even start about bonds! You own gold and bonds because you don't want to bet on stocks - right? It's like buying insurance against market downturns. I guess if you're not worried about downturns (i.e. if 2009 happened again you'd sleep just fine at night) you don't need them.

Anyway, I continue to be curious why people want to discount the gold runups of the 1970s and late 2000's, as the gold performance in those periods is exactly what one would expect given what was happening at the time: stagflation in the 1970s, and market uncertainty in the late 2000s. Looked at another way, if you think gold's performance in the 1970s was entirely due to relaxing restrictions on ownership, then you are saying that Harry Browne's view of the role of gold is completely wrong. If he made that big an error, then you probably don't want to get anywhere near the PP, as presumably that wouldn't have been his only mistake.

Also, I'm not sure the backtests about the returns of intermediate treasuries being a good enough substitute for gold are valid going forward. Rates were a lot higher in the past, and for long periods cash beat inflation - which it is not now, and may not for a long time. Incidentally, I also think that this is a big reason for the runup in stocks. People are buying stocks because the S&P 500 dividend yield beats CD interest rates, and gets better tax treatment. I wonder what's going to happen when interest rates rise and that's no longer the case.

Anyway just some (also over-caffeinated) thoughts....

[dualstow]

You and barrett drinking a lot of covfefe today, eh?

Anyway, I continue to be curious why people want to discount the gold runups of the 1970s and late 2000's

Well, you're so right, Sophie.
I would start by saying that I don't think Harry Browne made an error. There are smart people who like gold and smart people who don't like it. When you're playing with your own money, it leads to a lot more hand-wringing than merely reading a book and thinking, hmm, gold looks neat.

And while it's true that you could say the same about other assets, I suppose this has something to do with herd mentality. It's hard to shake the feeling that 10,000 Bogle fans can't be wrong. All those 60/40 portfolios that completely avoid gold and seem to be fine. (Yes, they might experience some pain in the future. One never knows). How many people have the converse, have gold and neither stocks nor bonds? Well, there's our Libertarian666. (sound of crickets)...and...a handful of people who aren't on the internet.

Let's say I did somebody wrong and my punishment is to sell everything and start a portfolio of pure treasuries. No problem. I could do that today, even knowing that inflation would probably ruin my plans. Same scenario, all stocks: hmm, a little bit harder, but I'd probably thank whoever's forcing me to do this in the long run. Like that movie in which Bette Midler's character thanks her captors for kidnapping her, because she finally lost weight exercising while chained up in their basement. All gold: yikes. I don't know.

(Could you answer my t-bill question now? :-)
Never mind, Tyler got it.)

[Cortopassi]

I only post this as informative in nature, please don't draw any conclusions...

One other aspect of gold, if held in physical form, not GLD at a broker, is it is effectively off the grid as much as you are willing to take it. I have yet to have any local coin dealer fill out forms when buying or selling gold.

[stuper1]

Exactly, so if there is a solar flare or electromagnetic weapon in the next war that knocks out the power for months, or some kind of financial shenanigans or hacker that erases everybody's brokerage accounts, you still have your physical gold. The calculations above show that you can hold as much as 20% physical gold and still get almost the same CAGR as somebody who holds no gold, plus it lowers the volatility of returns. I sure feel better holding physical gold and having insurance against the possible black swan events.

[Tyler]

Note: If one backtests starting in 1970, the conclusion will be that gold improves the portfolio significantly. If the backtesting starts in 1980, the opposite conclusion would be reached.

The way I look at it, you can always find a better or worse choice over a specific timeframe but gold improves the consistency of a portfolio across all economic environments.

[Tyler]

I suspect that's true, but I wonder how we can quantify it. I would think it would show up in the max withdrawal rate, which is essentially a measure of the consistency and level of CAGR.

Yes, SWRs are a pretty good indicator.

Lots of charts on the site illustrate it from different perspectives. But if you're looking for a single way to quantify it, play around with the Start Date Sensitivity calculator with and without gold. Here's a walkthrough that includes an example directly relevant to this conversation: https://portfoliocharts.com/2016/05/15/ ... an-others/

[sophie]

The problem with backtesting is that the results are going to be skewed depending on the prevalence of the different economic environments. If you started in 1980, you skipped past the inflation of the 1970s and are looking at mostly prosperity with a few years of deflation, plus a long, slow run-down in bond interest rates in the 1980s that could be regarded as an aftereffect of the 1970s inflation. So your backtesting won't be valid if another inflation comes along. Also, you could regard the 1980s hyper-prosperity as an artifact of post-inflation and thus not likely to be repeated. Bottom line, I don't trust backtests anymore.

It's true that there's not much you can turn to for advice on gold. Most writings are a mix of cherry-picked back-tests and highly biased opinion. Which is true of this little missive also, I guess :-). I arrived at my conclusions by examining the long history of gold and its function as an internationally and cross-culturally recognized store of value. I also spent time watching daily market movements of gold, long bonds, and stocks. Whenever stocks went down, either gold or long bonds (sometimes both) would go up. The opposite would happen when stocks went up. Despite the existence of other asset classes, gold and Treasury bonds are fundamental assets that are the main recipients of money that doesn't go into stocks or cash. Other asset classes can be regarded as linear combinations of the fundamental assets. For example, I tested corporate bonds against stocks and cash, and found that they track stocks so closely that I decided they're not worth holding as they provide no actual diversification.

There's also quite a lot of intelligent discussion about gold in the epic PP thread on the Bogleheads forum. It's worth reading through if you haven't already. Maybe it would be worth putting it into one long document and posting somewhere!

[Tyler]

Just to clear up a couple things: I actually really like gold. Somehow I tend to get on the anti-gold side of discussions around here at times, I think because I'm continually testing my own views on gold as an investment component. But I always end up choosing to continue to hold gold, even though I may not have the conviction to heartily recommend it to someone else.

Those who think critically of their investments and who question their own assumptions are the best types of investors! I always fully respect when a decision is well thought out and not simply an act of faith. Your points about the early performance of gold are all true and something worth considering. I just believe that even with those timeframes excluded portfolios with gold are appealing for their demonstrable consistency.

The problem with backtesting is that the results are going to be skewed depending on the prevalence of the different economic environments. If you started in 1980, you skipped past the inflation of the 1970s and are looking at mostly prosperity with a few years of deflation, plus a long, slow run-down in bond interest rates in the 1980s that could be regarded as an aftereffect of the 1970s inflation. So your backtesting won't be valid if another inflation comes along. Also, you could regard the 1980s hyper-prosperity as an artifact of post-inflation and thus not likely to be repeated. Bottom line, I don't trust backtests anymore.

IMHO, a wise investor should seek two pieces of information: 1) how a portfolio worked and will continue to work, and 2) evidence that it really did perform as advertised even in the least favorable timeframe. I think backtesting has developed a bad name because too many people cherry-pick data either intentionally or not and simply don't understand how uncertainty and sequence of returns works. But with the right mindset and the right tools, it doesn't have to be that way. I can't force people to think differently, but at least I can contribute to the tool chest.

[sophie]

Agree, backtesting helps you prove that a portfolio worked as it should during a given economic condition. I remember once toying around with inventing a way to backtest fairly by setting up a fictitious timeline with all economic conditions equally represented, using results from years in which each condition dominated. Or, set up multiple tests each focusing on one economic condition.

One thing that has always struck me is how every portfolio that doesn't hold gold got hammered in the 1970s. Since my goal is to preserve my assets even during the drawdown phase, I got very interested in this time period as the ultimate test of a portfolio's staying power in difficult long-term conditions. I tested to see if lowering gold to 10% would be enough to keep a portfolio afloat, and it's not. The minimum is just over 20%. If Desert is right and gold gains were artificially increased in the 1970s by relaxing of ownership rules, then you'd need even more. So it's not that I'm taking gold on blind faith.

Desert, great guess - it is indeed Marie Curie. Is your avatar Ben Franklin?

[Kbg]

I think Desert is exactly right. While the term bubble is probably overused, gold in 78-79 was absolutely bubbleiscious...way, way, way beyond any fundamental reasons.

[sophie]

Why couldn't there have been fundamental reasons?

In 1978, US inflation (December -> December CPI) was 9% (average 7.6%). The 3 month Treasury bill was paying 6-7%. In 1979, inflation went up to 13% by CPI (average 11%). The 3 month Treasury bill in that year spent most of the year at 9%, and went up to 12% by the end of the year. At the time, several European countries had much lower inflation rates.

That sounds to me like the exact scenario (rising inflation to > 10% with cash interest slow to catch up) that would cause investors to switch from buying Treasuries to gold. You're forgetting also that while gold may have been controlled in the US earlier in the decade, it was not in many other countries, and they collectively have a lot of power to influence the gold price with investment choices. The US was big at the time, but most certainly not the only player in the world gold market.

[barrett]

Why couldn't there have been fundamental reasons?

In 1978, US inflation (December -> December CPI) was 9% (average 7.6%). The 3 month Treasury bill was paying 6-7%. In 1979, inflation went up to 13% by CPI (average 11%). The 3 month Treasury bill in that year spent most of the year at 9%, and went up to 12% by the end of the year. At the time, several European countries had much lower inflation rates.

That sounds to me like the exact scenario (rising inflation to > 10% with cash interest slow to catch up) that would cause investors to switch from buying Treasuries to gold. You're forgetting also that while gold may have been controlled in the US earlier in the decade, it was not in many other countries, and they collectively have a lot of power to influence the gold price with investment choices. The US was big at the time, but most certainly not the only player in the world gold market.

Not to mention that the conventional 60/40 stock/bond split lost money in real terms for the 12-year period from 1970 to 1981. When financial assets like stocks and treasuries are doing poorly (because they are not keeping up with inflation), people have a tendency to buy hard assets like gold and housing. Housing prices in real terms kind of gyrated around in the 1970s but went consistently up during the years Sophie is citing above.

I totally get that the early 1970s gold data is questionable, but I believe that of the mid to late 1970s makes sense.

[sophie]

Barrett, that's a really good point: housing prices were subject to the same conditions in 1978-79 minus the history of price fixing. If they responded similarly to gold, that would be some evidence that the gold gains were not artificial.

Although, I personally can't believe that gold's fluctuations for the entire decade were all due to the effect of the previous price fixing. How do you explain the 25% drop in gold price in 1975? If 1973-4 were a gold bubble, I'd say a 25% drop is pretty effectively the end of it. That's about the same as the gold drop in 2013. I doubt anyone in 2014 was saying that gold was in a bubble.

What I'm more worried about with gold is the rise of ETFs, i.e. paper gold. It might be making gold more of a speculative than a safe haven asset. What exactly does a share in an ETF mean, if the managers can loan out a large proportion of their gold stores?

[mathjak107]

housing prices fell during the high inflation high interest rate years and did not recover until rates and inflation starting falling .

[barrett]

housing prices fell during the high inflation high interest rate years and did not recover until rates and inflation starting falling .

Source? From everything I can find housing went up in real dollar terms in the 1970s. The biggest rise in the late 1970s corresponded pretty closely with the rise in gold.

[D1984]

housing prices fell during the high inflation high interest rate years and did not recover until rates and inflation starting falling .

Source? From everything I can find housing went up in real dollar terms in the 1970s. The biggest rise in the late 1970s corresponded pretty closely with the rise in gold.

http://www.econ.yale.edu/~shiller/data/Fig3-1.xls

It gives both real and nominal values for a US residential housing price index; AFAIK this is a price only index and does not include what you would've made in rental income if you had rented the housing out.

For the period that stocks lost to inflation (1966-81 or 1969-81 depending on whether you go by the Dow or S&P 500) it shows nominal non-inflation adjusted residential housing prices from 1966-1981 increased just under 200% for 66-81 (i.e. if it was worth $10K on Jan 1st 1966 it was worth a little over $29.9K in December 1981) or approximately 177.94% for 69-81 (i.e. if it was worth $10K on Jan 1st 1969 it was worth around $27.8K in December 1981).

As per the Minneapolis Fed's inflation adjusted price calculator you would have had to have $24,790 in 1981 to buy what $10,000 would have bought in 1969; you would have had to have $28,000 in 1981 to buy what $10,000 would have bought in 1966. So it looks like for the whole late 60s to early 80s inflationary period residential housing beat inflation, albeit only slightly.

Now, for the 1978 to 1981 Volcker rate shock period, housing (as per Shiller's document referenced above), actually did slightly worse than inflation overall; the real inflation-adjusted value of residential housing went from approximately 113.4 at the beginning of 1978 to around 111.04 by December 1981.

[grapesofwrath]

I do like the 40 percent equity though, and am slowly building up from 30% to 40%.

You feel comfortable increasing your equity exposure at a time when US stocks are "richly" valued ?
I'm well below 30% and would like to increase towards 40% but am reluctant to do so now and wish to wait for a "mean reversion". Of course I could be waiting a long time as US stocks could be on a permanently high plateau.

[barrett]

housing prices fell during the high inflation high interest rate years and did not recover until rates and inflation starting falling .

Source? From everything I can find housing went up in real dollar terms in the 1970s. The biggest rise in the late 1970s corresponded pretty closely with the rise in gold.

http://www.econ.yale.edu/~shiller/data/Fig3-1.xls

It gives both real and nominal values for a US residential housing price index; AFAIK this is a price only index and does not include what you would've made in rental income if you had rented the housing out.

For the period that stocks lost to inflation (1966-81 or 1969-81 depending on whether you go by the Dow or S&P 500) it shows nominal non-inflation adjusted residential housing prices from 1966-1981 increased just under 200% for 66-81 (i.e. if it was worth $10K on Jan 1st 1966 it was worth a little over $29.9K in December 1981) or approximately 177.94% for 69-81 (i.e. if it was worth $10K on Jan 1st 1969 it was worth around $27.8K in December 1981).

As per the Minneapolis Fed's inflation adjusted price calculator you would have had to have $24,790 in 1981 to buy what $10,000 would have bought in 1969; you would have had to have $28,000 in 1981 to buy what $10,000 would have bought in 1966. So it looks like for the whole late 60s to early 80s inflationary period residential housing beat inflation, albeit only slightly.

Now, for the 1978 to 1981 Volcker rate shock period, housing (as per Shiller's document referenced above), actually did slightly worse than inflation overall; the real inflation-adjusted value of residential housing went from approximately 113.4 at the beginning of 1978 to around 111.04 by December 1981.

Here is one of the charts that I was looking at (this is inflation adjusted data as well):

https://inflationdata.com/articles/wp-c ... ousing.gif

I'll keep looking for better links but note that housing started to come down in real dollar terms in 1980 and continued to drop in 1981. 1981, as we all know, was a bad time to own just about anything. I believe T-Bills or money market funds were the relative "winners" that year.

Barrett, that's a really good point: housing prices were subject to the same conditions in 1978-79 minus the history of price fixing. If they responded similarly to gold, that would be some evidence that the gold gains were not artificial.

Although, I personally can't believe that gold's fluctuations for the entire decade were all due to the effect of the previous price fixing. How do you explain the 25% drop in gold price in 1975? If 1973-4 were a gold bubble, I'd say a 25% drop is pretty effectively the end of it. That's about the same as the gold drop in 2013. I doubt anyone in 2014 was saying that gold was in a bubble.

Sophie, also notice that the gold drop you are referring to corresponded with stocks shooting up. This gets back to Tyler's main point about gold... that it is largely uncorrelated with stocks.

[Kbg]

Why couldn't there have been fundamental reasons?

In 1978, US inflation (December -> December CPI) was 9% (average 7.6%). The 3 month Treasury bill was paying 6-7%. In 1979, inflation went up to 13% by CPI (average 11%). The 3 month Treasury bill in that year spent most of the year at 9%, and went up to 12% by the end of the year. At the time, several European countries had much lower inflation rates.

That sounds to me like the exact scenario (rising inflation to > 10% with cash interest slow to catch up) that would cause investors to switch from buying Treasuries to gold. You're forgetting also that while gold may have been controlled in the US earlier in the decade, it was not in many other countries, and they collectively have a lot of power to influence the gold price with investment choices. The US was big at the time, but most certainly not the only player in the world gold market.

Seriously...you don't think 1978-Jan 1980 wasn't a speculative bubble? I may give you 1978, but no way after that. Pretty much the definition of a bubble is up 400% in a year (79) and down by half the following year (80). Going near vertical in 2 months on a price chart is a pretty good tell as well.

[sophie]

Why couldn't there have been fundamental reasons?

In 1978, US inflation (December -> December CPI) was 9% (average 7.6%). The 3 month Treasury bill was paying 6-7%. In 1979, inflation went up to 13% by CPI (average 11%). The 3 month Treasury bill in that year spent most of the year at 9%, and went up to 12% by the end of the year. At the time, several European countries had much lower inflation rates.

That sounds to me like the exact scenario (rising inflation to > 10% with cash interest slow to catch up) that would cause investors to switch from buying Treasuries to gold. You're forgetting also that while gold may have been controlled in the US earlier in the decade, it was not in many other countries, and they collectively have a lot of power to influence the gold price with investment choices. The US was big at the time, but most certainly not the only player in the world gold market.

Seriously...you don't think 1978-Jan 1980 wasn't a speculative bubble? I may give you 1978, but no way after that. Pretty much the definition of a bubble is up 400% in a year (79) and down by half the following year (80). Going near vertical in 2 months on a price chart is a pretty good tell as well.

Nope. Why would a bubble be defined purely from price movements, ignoring the conditions that may have triggered it? And still less, why do you attribute the gold gains in this period to something that happened almost a decade earlier? There was a lot going on in 1980 that I don't have time to research at the moment - maybe you should?

Gold and treasuries rose after the 2008 financial shock as well. Treasuries rose ~30% in a very short time in 2008. Was that a "speculative bubble", or a logical result of what was going on in the markets at the time? Regardless of the term you use, those events are precisely why you hold gold and treasuries. People holding PP's in 2008 were calmly rebalancing and reaping the gains from these drastic price movements, while everyone else was in shock at watching their retirement account balances almost cut in half. I'm HAPPY to know that gold can move that fast in the right conditions. That's the whole point of owning it as part of a balanced portfolio.

[barrett]

Here's a good chart in nominal dollars (sorry about that!) so we know what we are debating:

http://goldprice.org/gold-price-chart.html

One of the things I remember about that time in late 1979/early 1980 is that the Hunt Brothers were attempting to corner the silver market. When they got over-leveraged, it caused a panic sell-off in silver that had a spillover effect into other commodities.

This link is not bad:

https://en.wikipedia.org/wiki/Silver_Thursday

I look at that spillover as pretty much the same thing that happens when the stock market gets slammed. Good companies are dragged down with the bad.

It's interesting to note that gold settled into a less crazy trading range through, say, the middle of the 1990s, and it did so at a price that was way above the value at the time of de-pegging. Maybe its price was in fact bubbly in 1979 but Sophie is right that that time was crazy. Among other things, cash yields were sky high.

Because of where gold "settled" after Volcker got the markets calmed down, I'm going to speculate that gold was really only in a bubble in very late 1979 to early 1980. Ditto for, say 2011 to early 2013.

Obviously we all have the benefit of hindsight now but another thing I remember about late 1979/early 1980 is that everyone was talking about the price of gold. Kind of like stocks in the late 1990s and housing in the middle of the aughts.

[Kbg]

Crazy is also associated with a bubble...curious as to what this group considers a bubble?

[l82start]

i always thought a bubble was self descriptive, a rise with nothing substantial underneath it..