I got involved in an interesting discussion about Bar Bell investing. The idea comes from splitting risk in bond futures trading or commodities trading. I traded commodities back in the 70’s and in fact saved up enough money I could either go to medical school or buy a seat on the Chicago Jackson street mini-market, so I was more than a little into it. Commodities trading is a zero sum game and a game of leverage. You might but a contract that has a potential for $1 gain but you may pay only 20 cents or 5 cents for the chance. Contracts expire so you can’t just buy and hold, your only choice if you’re in the game is to trade or get the hell out, so you place bets, and you win some and loose some. If you win more than you loose at the end of the year you have a profit, loose more than win a loss. Loose too much you’re busted. So it’s all about wining and its all about controlling your losses such that you have money to play. Commodities trading is not investing. It is trading on a market with the idea of buying low and selling high and the very real possibility of buying high and being forced to sell low. If you loose, what you loose goes to whomever had the other side of the trade your “buy high sell low” pair is his “buy low sell high” pair. If you don’t like that much risk you can do something called spread trades and there are all kinds of spread type trades.
So in a poker game you may have the main game going on and bets on the side. Bets on the side have a different risk of paying off than the main game. By knowledge of the odds of the side bets and the odds of the main game you can improve your risk by making enough on the side to make up to some extent for the risk of the main game. FIRE types do this all the time in fact they plan this into their strategy. They work 10 years with the idea of having that 10 year “investment” generate 60 years worth of income. Sounds psychotic doesn’t it? So they hedge their bets with a side bet called a “side gig”. A little blog or something that generates some income to cover part of the risk of the nutzo 10 years work 60 year play narrative. FIRE types think they are investing but actually they are speculating same as a commodities trader or a poker player. Small businesses fail 20% in the first year 30% in the second year and fully 50% fail in the first 5 years. 30% fail AFTER 10 years. So that little blog is anything but a sure thing. It is the reason startups sell out. Start up, make some money, sell out before you fail. Speculation. So what about Bar Bells?
The Bar Bell portfolio comes from splitting risk in bond trading. You can buy low yeild high quality paper and are pretty much assured of making your coupon. You can buy high yield low quality paper that if it pays off, pays off big but there is a definite risk of default. This is called a spread. You spread your risk between a sure bet which pays you and a speculation which may or may not pay you, and this is why it’s called a Bar Bell. It’s a bi modal risk. Pictorially
Here is a balanced portfolio of known risk and known reward and it sits on the efficient frontier. It has an expected return, the peak at the center, and an expected risk the range of values above +- 3. What if you squish down the center?
You get a bimodal distribution with a risky end (pink) and a safe end (blue). This is a Bar Bell The argument is you split your risk between safe and risky and if risky dumps you still have safe but in reality you have just the same amount of area under the curve in Bell as Bi-Modal it’s just distributed differently. The net portfolio will still work out to some average risk and reward. What if you have a BIG blue and a LITTLE pink but pink has a chance to multiply dramatically or it can go away completely? Lets look at some charts
Here is a portfolio of SPY. You retire in Dec 1999 just before the 2000 .com crash and also experience 2008. You’re a Bogglehead and expect SPY to generate 8.82% return over the long term.
Spy between 2000 and 2018 under performs BY A LOT
You’re expecting 8.82%, your actually getting 4.58% or half of what you expected. SPY experienced 2 recessions in that 18 years and so the volatility ate your lunch and your nearly 2/3 done with retirement! How much “return is it going to take to get you back to your expected 8.82% over the next 12 years? If you started with 1M and compounded at 4.58 x 18 years you would have 2.4M Your expected at 30 years with 8.82% is 12.6M, so at 18 years you are 10.2M under expected and NEED to make 13% / year for the next 12 years to make up the difference. The expected is 8.82% Do you really think it’s going to work out? THIS IS SEQUENCE OF RETURN and this is a real present day example not just some speculation This is exactly what happened to a guy who retired in Dec 1999 with 100% SPY
Let’s Monte Carlo a SPY portfolio for 30 years 4% WR
over 1/10 times you 100% run out of money. You run out of money because of the RISK in SPY
Here is the kicker to 100% SPY
Your portfolio starts dying EARLY. By 15 years it’s already heading down and for some is out of money. Retire at 60 get to 75 Uber Driver here you come!
Lets say you buy only VBMFX which was has been around since 2000. It’s expected return is
almost 5% and in fact it did better than that. Over the past 18 years VBMFX has returned
5.75% since inception in 2000. Bonds have been in a 30 year bull market.
So if you had SPY and took out 4% you could expect 0.6% growth on your money, less than inflation in the past 18 years, where as if you had VBMFX you could expect 1.75% growth about equal to inflation. What happens if you Monte Carlo VBMFX?
You survive 98% of the time but notice purple is headed into the dirt.
What about the kicker, what about the kicker??
The kicker is you don’t start running out of money till after you’re probably dead with this SORR
SPY alone and VBMFX alone are Bi-Modal when you look at them together, and they are almost perfectly uncorrelated as we saw in the example. Stocks under performed Bonds over performed. What happens if you Model a 50/50?
You fail about 2.8% of the time but look at purple, it is barely headed into the dirt and will survive for many years to come. The idea is to then have a “less risky part of the portfolio and a risky part for a Bell so lets add 20% AMZN for a 20/30/50 AMZN/SPY/VBMFX portfolio
Holy Cow! a >99% survival!
All you needed to do was to know to buy AMZN in 2000
Substitute GE of AMZN for a 20/30/50 GE/SPY/VBMFX portfolio Uh oh 89% survival no better than SPY!
It can be even more complicated if you change SORR or Inflation assumptions.
So what does all this mean? By choosing at Bar Bell you are speculating just the same as a poker player with side bets. Some times you win some times you loose, good time Charlie has the blues! How do you apply risk management? 1. You over fund to start. If you start at 20% above your 100% of need it can be a home run or a strikeout. If you’re a 3%/33 type with a million bucks you need 1.2M to start. An extra 200K is roughly an extra 7 years of work. If you strike out, YOU’RE OUT. It means you weren’t cut out to be a speculator. You don’t know what you are doing. Do not put hamburger money in the pot. See #2.
2. NEVER send money from low risk to high risk. You may send money the other way, in fact that’s how you get rich in this kind of scheme, dollar cost average the risk. When things are up stuff some into the mattress When things are down you have your 100% of need so go to the beach. Another way to play this is to completely divorce the portfolios into 2 independent portfolios one for investing and one for speculation. WR is strictly NOT part of speculation. The flow sheet for this kind of risk management is
You are always siphoning profit or principal off into investment never the other way. Here is my story with this technique but not with these numbers. The ratios are correct. In 2005 I bought a penny stock for 10K. It grew to 50K and I pulled out my 10K and put that in BRK.B. This is called a free trade. The remaining 40K was free money. The investment bobbled around till 2015 when I sold that 40K to cash. When I sold it it was worth 50K. I put half in BRK.B and half into BTC. 25K into speculation and 25K into investing. I had made 500% profit, and my principal was tucked in BRK.B plus half my profit. BTC exploded and my 25K went up to 1.5M. On the way up I took out the 25K and put it into BRK.B (free trade) and left the free money alone. BTC crashed OH WOE IS ME? Hell no The money was free money. It could go to zero and I’d be out nothing. As it turns out it’s still up 1300% and I have all that loot in BRK.B to boot. If I went to zero Whoope shit I had a good ride and made some money which I stuffed into BRK.B. But the thing is BTC is not going to zero. BTC is creative destruction and creative destruction multiples. AMZN is creative destruction. Apple smart phone is creative destruction. NFLX is creative destruction. Remember the brick and mortar Block Buster? NOT creative destruction. In the mean time I’m going to the beach quite content to do nothing.
The point is if you had co-mingled the portfolios the risk of one would seep into and erode the value of your hamburger money. In retirement do not let greed eat your hamburger money. Do not be a Dumb Bell because you read a Bogglehead book and think you know something. Reading a Bogglehead book doesn’t mean you know Jack it means you know Taylor Larimore, who’s just an old coot with a system (sorry couldn’t resist). These traders make their living eating your lunch and they are damn good at it. This ain’t bean bag. If you can’t practice risk management, don’t play. If you don’t know what you’re doing, don’t play. Speculation is a zero sum game. Some’s gonna win some’s gonna loose.