A novel approach to portfolio construction...

The traditional approach to building an investment portfolio with the aim of achieving a target wealth level for a set date in the future (typically retirement date) is relatively straight forward and involves discounting the future value of this portfolio to the present using an estimated discount rate which comprises a somewhat rough estimate of the expected annual growth rate of the portfolio. The major flaw with this approach is that the rate of return of the portfolio is very difficult to estimate, especially in the current volatile environment where visibility is relatively poor.

A far more intuitive and robust proposal would involve splitting an investment portfolio into two segments. One segment would be designated as the exposure that aims for capital preservation. This exposure contains investments that have very low risk and very low probability of experiencing drawdowns. The size of this segment should reflect the absolute strict minimum amount of wealth that the investor will need once he or she reaches retirement. The other segment comprises of the significantly riskier portion of the portfolio that the investors could afford to lose entirely but that should provide steady growth of the capital over the longer term.

You may wonder what the difference between this approach and a classic asset allocation portfolio comprised of a diversification between bonds and equities may be. The key difference is the the so called low risk allocation is there purely to preserve capital, it won't grow on its own but will increase in time through the reallocation of capital from the riskier segment, assuming that this segment does actually grow. The classical approach runs the risk of experiencing a sharp increase in correlations similar to 2008 when practically all asset classes were losing value simultaneously. Drawdowns in this novel approach should, instead, be limited to the "risk" portion of the portfolio.

The challenge in implementing this is twofold:

1. How do you ensure that the "capital preservation" allocation does its job in a world where capital losses may be incurred even with money market deposits, through the uncertainty regarding counterparty risk (Lehman's being the recent example)
2. The capital transfer from high risk to the low risk segments assumes that the high risk segment will grow through time. Nothing is less certain if we look at stock market performances over the more recent past!

Still, despite these challenges, the novel approach does provide a solution that is far more adapted to the current market environment and thus has a better chance of succeeding over the longer term.

Ripple effects...

A deal was struck (if we can call it a deal) last week between Greece and its private bondholders where up to 80% of them "agreed" to exchange their toxic debt against new ones that are thought to be less than half the face value. I guess the rational behind the acceptance was that it was a choice between getting something or nothing.

The more interesting question to ask is if this deal will trigger payments from credit default swap insurance on Greek bonds. Normally it should because the agreement is clearly akin to a technical default. If the body that decides on such matters argues that it is not a default, they run the risk of significantly damaging the credibility of the CDS markets, causing even greater burden on the borrowing of the other Euro-zone economies that are struggling with their debt.

If, on the other hand, they consider this action as a default, there is still great uncertainty as to how it will unfold. Just as in the Lehman and Dexia debacles in 2008, nobody really knows what the counterparty risk really is. We may have an idea of the amount of exposure and the main counterparties but we don't, for example, really know the counterparty to those counterparties.

The point is that in either of the two scenarios, there is bound to be ripple effects and it is difficult to estimate beforehand what the outcome of those ripple effects are likely to be.

If reason prevails (not always a given in financial markets), the governing body will rightly recognize the deal as the equivalent to a default. Greece has evidently defaulted on its debt, the question is, where will it spread next?

What is an equity index really measuring?

Equity indices are designed to measure the performance of a specific sector, market, region or even the entire world. They are meant to provide an indication of the type of returns you can "expect" from equities over the longer term but they also tend to suffer from some serious shortcomings that can have serious consequences.

Most popular indices for example only capture the price performance of its constituents typically weighted by capitalization (S&P 500) or just price (Dow Jones Industrials). What they don't capture, however, is the impact of dividends, especially when they are reinvested. Total return indices do exist, but they are unfortunately much less common and therefore rarely used and so investors are much less aware of them.

Even more rare are indices that are adjusted for inflation. That type of information is just not published but the impact can be very significant.

To show just how significant the effect of dividends and inflation can be on the return of an index, researchers in the U.S. did the calculations on the Dow Jones Industrials index and here is what they discovered:

If dividends were included since the inception of the DJI in 1896, it would have a value of more than 1,300,000 which is more than 100 times its recent high of 13,000!

If it was adjusted for inflation over the same period, it would shrink to a value today of around 500 which his about 96% below its actual current value.

Finally, if we were to combine the impact of reinvested dividends and inflation, the figure would be around 47,000 or almost 4 times its current value! In other words, the impact of reinvested dividends seems to be substantially greater than the impact of inflation (assuming inflation is correctly measured of course!)

All this to say that it is important to be aware of what exactly is being measured (or not) when looking at an index. Just as in the differences between price weighted and capitalization weighted indices, the effect of including dividends and/or inflation can lead to substantial differences in the results.