I woke up this morning thinking about diversity, as in non-correlated diversity. There is correlated diversity and a vehicle like the S&P 500 is all about correlated diversity. The stock mix reduces single stock risk down to market risk, but once market risk is achieved there is not much to be gained in Piling more and more issues higher and Deeper (PhD). 20 stocks across 10 sectors is 95% diverse. Adding 980 more stocks only makes you 99% diverse. A nod is as good as a wink to a dead horse. I study the S&P 500 and all year long it’s been troubled. 20% of the stocks were in a bear market, 20% more were under water, meaning 200 of 500 stocks were doing bad the entire year despite the “RAH RAH best economy in history”. Only 40 stocks were responsible for the majority of the gains and especially the FANG plus a couple, at any time. The index has broken down and those 40 are no longer performing and some of the high FANG flyers are now in a bear. How is this diversity when you are relying on 40 stocks to give you gain? The problem is these stocks are highly but not perfectly correlated and so stocks have a general direction when going up and a very tight path when headed into the dirt, like a hand. On the way up stocks look like an open hand each finger pointing in a different direction slightly diversified and slightly reducing risk. On the way down the hand becomes a fist plowing into the ground. The algorithms sell first and ask questions later. Because of the high correlation the in the bad times algorithms erase the “illusion of diversity”.
Here is a picture of GLD (gold line) v S&P500 (blue line) since 2013. The correlation between GLD and S&P 500 is 0.04, and the picture shows that gold is flat while S&P is exploding. This is non-correlated diversity, the kind of diversity that saves you.
This is a longer term picture of S&P v GLD including the 2008 debacle. Looking at GLD from 2005 when the S&P went in the toilet and dropped 50% GLD exploded in value, the typical flight to quality. Again you see the benefit of non-correlation. Stocks went down and then stayed flat for a long time while GLD went way up.
Stocks are property, GLD is property, just as Real Estate is property and Bonds are property. These assets are not money, they however can be converted into money using a market mechanism. The property is worth whatever someone is willing to give you for it at the instant you want or need to sell it. The way you make money is buy the property low sell the property high. The way you loose money is buy property high sell the stuff low. Suppose you retired in 2005 and are living off your assets. What asset would you sell in 2008? You bought GLD low sometime previous 2005 and you bought S&P high sometime previous to 2008. Buy low Sell high, you would sell some GLD (sell high) and keep the stocks or even add to the stocks (buy low). You sell some GLD and have money for hamburgers!! Selling the stocks low would blow up your compounding plan in the long run. If all you own is stocks you’re hosed.
Let’s add some Bonds (BND the teal line) In a 3 asset portfolio BND and S&P has a .05 correlation. GLD and BND has a .44 correlation so BND pretty much does its own thing as the chart shows. BND remains flat while S&P is crashing and GLD is soaring. In the case of a stock crash, BND might be a good source of hamburgers, or maybe a lil’ GLD AND a lil’ BND. You see you now have 2 choices that won’t cream your retirement plan by being forced to sell stocks low. Your GLD doesn’t have to last you forever only long enough to get through the bad time. You can always buy some more from William Devane when the price comes down later. It gives a ready source of value and gets you a year closer to death with your growth motor (stocks) intact.
You say yea but I’m a Real Estate guy!!! So let’s add some REIT
Purple (VNQ) doesn’t impress. It doesn’t grow well but dives into the ground just fine, even more so than stocks and in the same time frame! The correlation between stocks and VNQ is .72 You say I’m a Globalist! I have Global to save my butt! It’s diversity I tell ya! Let’s add VHGEX which is .96 correlated with S&P.
Salmon (VHGEX) looks pretty much like dark blue to me from a diversity perspective. If you gotta sell something you’ll be selling salmon low same as dark blue in a crash. This is the story of non-correlated diversity.
I was thinking about dividends. In a crash dividends at least for a while tend not to change even though the asset value plummets, so dividends may be a source of diversity but I’m not sure how to model that. Many people brag about living on dividends as if that’s safe. Not so sure it’s safe but the diversity may pay you.
Here is a calculator that looks at S&P 500 since Dec 1999. I chose Dec 1999 since it would represent 18 years into a retirement spanning 2 downturns and is familiar since most of us lived it. The average return on the S&P in this 18 yr period is 3.8% (NOT 10%). The reinvested growth is 5.8% The dividend is 2%. When you spend the dividend it is not really different than selling stocks, you just “sell” before you even “buy” (re-invest) The fact the dividend might be considered diversified hurts and helps you. You have a relatively stable stream at least for a while in a crash, but not re-investing means you don’t use that money to “buy low” in a crash.
Here is the inflation adjusted dope:
Inflation adjusted the S&P 500 sans dividends over the past 18 years has only grown 1.6% per year and only 3.5% with dividends reinvested. Dividends therefore were 1.9% inflation adjusted. To me this says dividends pay because of their non-correlation not because they are “safe”. Next time some bogglehead putz tells you it’s stupid to own GLD tell em to go pound sand!