Tuesday, June 23, 2009

Correlation revealed

So now that my secret is exposed that I try to mimic the entire structure of the capital markets it is best to look back at how the portfolio would have performed. Here we arrive at some very tricky parlance and one must be very careful when listening to investment advisers when they talk about performance. See some will talk about relative performance and some will talk about absolute performance and a non-mutually exclusive group will talk about one or the other spinning the performance in the most positive light. More on this in a moment.

In an earlier posting I proclaimed I had no benchmark, which is true. To offer a benchmark I would need someone mimicking my portfolio. So I have no S&P500 guiding light. However, I still like to use it not as a benchmark but as a fixed position to compare where I have been and where I am headed.

So I took the portfolio and how it performed in 2006, 2007, 2008 and the 1st half of 2009. I compared it to a buy and hold versus a rebalanced portfolio and then I looked at the S&P500. Again, I cannot state that the S&P500 is a benchmark, but I find it is as the most useful position line in which to compare. This is because, it is by far the easiest and probably wisest investment that a novice investor could undertake. One would need to surf to Vanguard's website and just choose the ETF or mutual fund that mimics the index. S/he will probably be in the top 1% of his investment cohort at the end of ten years. This is because s/he will pay minimal fees and not pay a manager for alpha generation, including failed alpha generation.

I have attached the spreadsheet here. All the information in it is from Yahoo Finance.

The first sheet has the funds and their respective performances. There is also a table for the S&P500. The second more interesting sheet is the Portfolio performance.

The first table is the ETF's performance again. BND and ESD both have their monthly dividend yield incorporated because it is a major part of their returns. The rest are just their price performance. One can see that they are all over the map. Some are well correlated, others less so. You can also find covariances and create a covariance of the portfolio. Then you could find sigma or the variance of the portfolio and produce a hypothetical ultimate portfolio. I have that somewhere based on these same numbers. However, as you may have guessed with such limited numbers this ultimate portfolio is of limited use. One of my main arguments against it is that the correlations change depending on how you frame them.

An example, gold is traditionally a hedge for your portfolio. In bad times, gold retains its value as a precious metal, as a unit of currency. So its value is as an insurance policy against your portfolio dropping. However, if you compare its performance to VWO, the most likely candidate for the highest beta in the portfolio, they were almost perfectly correlated (small sample size) until 2008. Then they became highly uncorrelated. So the perfect portfolio I find is largely a myth. One would need to keep generating a new one over and over as more information is incorporated into the secondary market exchanges. This ultimately only benefits your broker as you keep transitioning in and out of asset classes chasing the unicorn from Optimal Portfolio.

I choose the asset classes that fit best with my investment thesis.

The thesis in brief:
  • equities are better than bonds in the long run. So I want a significant portion in domestic and foreign markets to hedge against inflation. All central banks aim implicitly or explicitly for 2% inflation per year. This in turn will take away a lot of your fixed income stream in the long run. If you bought a 30 year bond offering coupons of 60 dollars a year the 60 dollar annual coupon is worth around 49 dollars in 10 years, 40 in 20 years and the last coupon is worth 33 dollars. You need to have income in real terms and equities are the best way to be exposed to real income. So I want about 35% of my portfolio there
  • bonds are necessary because they are more of a sure bet than equities. They offer stodgy returns but you can count on the principal repayment and coupons less the default risk. However, they offer only nominal payments that are eroded by central banks tacit inflation benchmark. So only 20% here.
  • 45% to real assets. In a fiat world where money can be manipulated by central banks and legislatures the only real income one can count on is real assets. From commodities to precious metals to real estate, I want stores of value. I never expect the governments to allow deflation, even in our current balance sheet recession. Even as people, governments and corporations unwind their debt, deflating asset prices, I believe in the Federal Reserve and their brethren to produce their low targeted inflation. Thus the large exposure to this asset class
  • Finally, as David Swensen of the Yale Endowment makes clear the difference between the top performing bond manager and the bottom performing bond manager is about 1%, 100 bps. About a 3% channel in equities. So there is no real gain in spending an inordinate amount of time searching for yield by picking the best manager. In fact today's best may be tomorrow's dog. Near everyone can recite the lore that the sun shines on a dog's ass some time. So I index.
  • If at some time the financial markets become evolved enough to allow ordinary investors the crack at venture capital and private equity managers, I would definitely be interested but the questions and concerns for these titans and their investment vehicles are best saved for when that day is closer at hand.
So after three years what do we have. A big mess. On a relative basis (your ears should be perking up) my two portfolios the buy and hold and the rebalanced both outperformed the S&P500 by a considerable amount. As in over 900 bps or 9% better per annum. As an absolute performance the portfolios both lost over 2% a year to finish lower than the initial outlay. That is, during this time period it would have been wiser to have been in Treasuries or Bank CD's. That is a depressing mess.

The buy and hold did perform slightly better than the rebalanced portfolio but they are really neck and neck. It will be better comparison when I review this portfolio again at the close of 2009.

So at the end all this babble about increasing your return. I only lost 800 of your 10,000 dollars whereas the S&P500 would have lost you over 3,000 dollars. So not a real lot to toot my horn on. However, this portfolio is a long run portfolio and I have more tricks up my sleeve.

Over the next few weeks I will show how you could have avoided all this loss and instead been up by close to 70% over this same time period with these same investments. However, I still believe that in the long haul, over the next set of 5 and 10 year periods this will be a huge money maker for your retirement and that you will be a top your cohort of newly minted investors following this simple and painless strategy.

No comments:

Post a Comment