Larry Swedroe's "Reducing the Risk of Black Swans"

[mathjak107]

See?! You couldn't "stay the course" either!

It would be interesting to see a FIRECalc-type run of the "classic" 4 assett Harry Browne portfolio vs a more conventional 60S-40B portfolio with 3% to 4% withdrawals and the long historical data set (i.e. not starting in the 1970s). How would median available spending rates compare over 30 year timeframes using the "4% of year end portfolio value" withdrawal method? While gold has sometimes "comes through" to help portfolio values, it doesn't always do it at the right time, and overall it is a considerable drag on overall performance (when compared to the stocks it displaces--those dividends can be darn handy and gold ain't paying any).

IMO, the strongest case for the PP is in a real disaster (international depression, bank system failure, loss in confidence in the currency, hyperinflation, etc). A real "black swan" in the classic sense--an event that most people agree is impossible/extremely unlikely. And even Harry Browne's classic portfolio assumes US Treasuries will stay "solid", something that the more hard-core "financial preppers" would not agree with today. Harry liked Swiss instruments back in the day--I wonder what he'd think of the more recent Swiss proclivity to link/delink their currency with the euro. And why exactly 25% to each "pot"--seems suspiciously like TLAR*. Maybe 10% - 15% gold is enough to cover contingencies--where did 25% come from? Or maybe Harry was just being honest: TLAR is the best any of us can do, implying more precision in our AA given the real data set and huge assumptions may be just kidding ourselves.

Protecting against "black swans" may reduce a portfolio's ability to perform well in normal circumstances--the ones most likely to happen.

*TLAR: That Looks About Right

i did it for about 10 years but being in my prime investing years i wanted more gusto.

 

[NW-Bound]

Protecting against "black swans" may reduce a portfolio's ability to perform well in normal circumstances--the ones most likely to happen.

One cannot buy insurance, then complain about the cost of the premium.

 

[samclem]

The difference - and this is what's missing in the Ferri thread - is that folks on this board are talking about a 30% Total Stock Market or S&P allocation while Swedroe's talking about only having small value, international small value and emerging markets in that 30% equity calculation.

I'm a believer in the small/value premium, so much so that I've tilted my portfolio that way. Value especially is very attractive to me due to the higher dividend yield and the (apparent) ability to benefit from the over-reaction of crowds to various trends and bits of news. I admit I haven't read Ferri's new short book, but I am concerned that there's a bit of datamining going on here--these sub-niches Ferri has seized on have done well under a particular set of market conditions, and there's considerable dispute (ever since Fama & French wrote their seminal paper) on why the effect exists and whether it will continue. So, while I believe it has worked and have structured my portfolio to take advantage of it, I would not consider it worth hanging my hat on so I own large & growth, too. I should try to discern why Ferri thinks the outperformance of these niche-ish subsectors is robust enough to have folks betting their future on. Because if they fail to do very well in the future, the large (70%) allocation to assets that can be expected to barely match inflation will leave many retirees short of money.

 

[kevink]

It'll take you even longer to find the answers samclem if you go looking for it in Ferri's [non-existent] book, but the one by Larry Swedroe does have a whole chapter that addresses your (and my!) concerns about whether there's data mining going on, whether his strongly tilted allocations are sufficiently diversified, etc.

All good questions on your part, that's for sure. Thanks for adding so much to this thread.

 

[samclem]

It'll take you even longer to find the answers samclem if you go looking for it in Ferri's [non-existent] book

Uggh, thanks. Heck, I would like to see what Ferri thinks of Swedrow's "deliberately undiversified equities to enable reduced overall portfolio risk" idea. Pfau and Kitces, too.

 

[kevink]

Me too! I know what John Bogle would say!

 

[mathjak107]

more and more i am thinking of splitting my portfolio money in half . half will stay in my growth and income model but i am seriously giving thought to going back to using the basic 4 part permanent portfolio.

with such a poor outlook for things at this point and a 7 year bull market i can see a flight to safety causing those long term treasury's to reverse direction in a heart beat and soar . .

the cash is like a call option on cheaper prices on whatever asset craps the bed and gold has fallen enough to be worth holding again.

sure it is easy to compare other portfolio's to find better combo's at different time frames in the past but no one can find that better portfolio for going forward.

i think i would find a mix of the two portfolio's quite comforting.

 

[RobLJ]

I have shifted toward foreign/dividend, small/value as 40% of the equity allocation, influenced by Swerdloe and others. It was crushed last year by the S&P but this year is outperforming.
While I like the PP idea (I bought it in '06 with DW's bonus money for the kid's college money because I sniffed a crash coming), I currently don't like the long bond piece. In a crash I would be wrong, but longer term, I just don't see the risk/reward. At all. I shifted the bonds over the last 3 years to intermediate and foreign/corporate and cash (for buying stocks in a significant correction).
It's timing, but once the Fed raises rates, I will start cost averaging into LT.
I'd be tempted to buy options on a market decline as opposed to long bonds, as insurance, but only in this environment, but I won't do it, since virtually all of my money is in 403b/401ks.

 

[mathjak107]

the long bonds play a very important part. in a down turn they usually soar. in 2008 TLT was up around 35% and the permanent portfolio was up that year. corporate bonds were barely able to stay positive under the same conditions forget about doing any heavy lifting . using options not only do you have to guess which direction things are going but when they will do it and even by how much and if you are wrong you lose money.

so the long bonds are your fighter cover and do not really come in to play until the enemy attacks. when markets are up they usually are down .

the permanent portfolio needed to do little the last 6 years or so since it never fell thanks to the long bonds.

while long bonds were being shunned just prior they were the winner in the end.

all the parts play a crucial roll , even the cash heavy position is actually acting like owning call options on assets at lower prices . how cool would it have been holding 25% cash in 2008-2009 when markets fell 6000 points .

you can't really peel out the individual parts and aye or nay them as working together they are better investments than anyone of them standing alone.

do you know if you had the worst timing ever and decided to buy the 4 part permanent portfolio at the moment gold peaked back in the 1980's
that just from rebalancing over all that time your gold would actually have beaten your return on the s&p 500 part just before gold rolled back more recently.

if i remember golds average was 9.8% vs the s&p 500's 9.6% at the time . that was while buying your gold at the worst moment in history.

yes , the cycles for the PP are longer than typical but over time each asset gets to have their day in the sun and then winning happens by not losing.

the 15 year average for the PP i think is still about the same if i remember as a 100% investment in the s&p 500 .

i don't mean the PP fund either , as that has gotten to far away from the basic 4 part portfolio as well as it being to expensive .

i went back to the PP just this week and had to decide whether to chance timing buying the parts. i realized that the whole purpose of going back to it was the uncertainty of things so trying to guess and time the purchases is not something i wanted to do.


so now i have half the money in the PP and the other half invested normally in the fidelity insight growth and income model.

basically i replaced the insight income and capital preservation model i was using with the PP which was not just a bet on low rates and good market days.

 

[mathjak107]

Nice charts to look at, but they hide a lot of data due to the ">" and "<" in the color key. This helps the low volatility PP look better (lots of green!) and doesn't give sufficient "credit" to the high volatility, high return TSM portfolio (e.g. from 2012 through 2014, the VGD Total Stock Market had a total return of 72%, or approx 20% CAGR. The chart just shows it did "better than 9%" Yeah--LOTS better).

Tyler's charts go out just 10 years. The PP has just 25% stock, the rest is in bonds and gold. If that mix is going to lose ground to inflation compared to the more conventional 40-60% stock, 40-60% bond portfolio, it might take longer than 10 years.

I think what most ERs want is enough sustainable long-term growth to stay ahead of inflation. Low volatility is a plus, but if it comes at the expense of a greater likelihood of losing out to inflation over 20-40 year periods, then somewhat higher year-to-year volatility is preferable. In my opinion . . .

what is really missing in the back testing of the swr of the permanent portfolio is the fact that all data starts in the 1970's while the worst case scenario's for retirees were 1907 ,1929, 1937 and 1965-66.

there is nothing to back test with since gold was fixed at 35 an ounce prior.

silver is not a good proxy for gold either since silver plunged in 2008 while gold was flat to up a bit so they are not always joined at the hip .


i am not saying it wouldn't have held up well , just the fact that charts produced showing swr rates starting in the 1970's missed the worst time frames the 4% swr is based on and there is no way to know what they would have done since a major component has no history..

 

[Ed_The_Gypsy]

Thanks for the Rick Ferri article! Looking at his charts makes me think of Wellington for accumulation phase (60/40) and Wellesley (40/60) for de-cumulation with expectations of no more than 5% returns going foreward.

Or, back to 100-age = bonds%.

The common theme throughout is the old one of trading potential higher return and its volatility for lower, more stable returns.

Sent from my SM-G900V using Early Retirement Forum mobile app

 

[samclem]

there is nothing to back test with since gold was fixed at 35 an ounce prior.

The Breton Woods arrangement was comprehensive, but would it be possible to derive a fairly accurate USD/oz value for gold pre 1972 by finding someplace where it was openly traded and it's value fluctuated in local cuurency (Switzerland?) and then convert their price to USD using the published exchange rate for their currency? This wouldn't be perfect (because the fact that gold wasn't publicly available for trading in much of the developed world surely affected its value), but it might be better than nothing.

 

[mathjak107]

with the dow down 220 today the permanent portfolio is actually up.

tlt up over 2% at the moment and gld up .50% .

this is just how it usually works with the PP MAKING IT WHILE ONE ASSET CLASS IS LOSING IT on those very volatile days.