Courtesy of Business Times

Published February 2, 2009
Why the Dow is holding at 8,000
By R SIVANITHY

TO MOST casual observers, the fact that the Dow Jones Industrial Average (DJIA) has bounced back every time it dipped below 8,000 points over the past few months - even when there is bad news - suggests that the 8,000 mark is where the 'support' or the magical 'market bottom' lies. This means that as soon as the index nears 8,000 on the downside, chartists and traders will start calling a 'buy' on the market.

The divisor: For every dollar an index stock falls, the DJIA falls 7.964782 points, regardless of the stock's capitalisation

Closer examination, however, reveals that the bounces around the 8,000 mark are simply a function of the way the index is constructed. Because the Dow is price-weighted, it is also inherently flawed.

In Thoughts from the Frontline weekly newsletter dated Jan 23, writer John Mauldin correctly points out that the divisor for the DJIA is 7.964782, which means that for every dollar an index stock falls, the DJIA falls 7.964782 points, regardless of the stock's capitalisation.

As a result, if the stock of Microsoft, with a price of US$17 and a market cap of US$156 billion, was to crash to zero, the DJIA would only lose 135 points (17x7.964782). But if the same was to happen to IBM, with a smaller market cap of US$124 billion but a higher share price of US$92, it would cost the index to lose a whopping 700 points.

Now consider the four financial stocks currently in the DJIA - Citigroup (US$3.90), Bank of America (US$6.78) Amex (US$16.70) and JPMorgan (US$25.43) - using last Thursday's prices.
If all four stocks were to crash to zero, the DJIA would only lose 300 plus points, not that huge a loss in the context of the market, yet imagine the repercussions on the US and global economies if these four institutions collapsed totally.

Most of the news on Wall Street these days centres on the crippled financial and auto sectors. But because the share prices of these companies are now so low, these stocks do not affect the DJIA by much (General Motors' shares, for example, are now just above US$3).

In other words, because the index stocks most affected by bad news are already battered to rock-bottom levels, the DJIA doesn't seem to fall much when bad news is released, thus giving the mistaken impression of resilience to adverse news and of strong support around 8,000 points.

By right, these financial and auto stocks should have been removed from the index, given that it has been past practice to replace stocks whose prices drop below US$10.

For some reason, the DJIA's guardians have been reluctant to do the same now, possibly because of the political fallout that might ensue - imagine the repercussions of removing pillars like Citigroup or General Motors.

This then leads to the inevitable conclusions: the DJIA is not comparable over time; the only reason the DJIA appears well-supported around 8,000 is because the collapsed financial and auto components have not been replaced as they should have been; and that movements in large-price stocks are magnified because the index is heavily skewed in favour of these counters.
If the index was to be correctly re-balanced by removing the battered financials and autos and replacing them with stocks with prices above US$10, you'd have to wonder whether the 8,000 mark would hold as well as it has.

You'd also have to dismiss arguments that it is safe to buy since the index is at its lowest level in many years because historical comparisons are invalid - unless, of course, the same re-balancings that were done in the past are performed now.

How to overcome such a large distortion? The most commonly accepted solution is to use market-cap weights, but this too has its drawbacks.

Last Thursday, the market-cap weighted Straits Times Index (STI) rose 0.64 of a point to 1,766.72, a move that a casual observer might interpret to indicate a mixed or quiet session. Far from it - if you stripped out warrants, the rest of the market only recorded 95 rises against 188 falls, a gain/loss ratio that indicated market weakness rather than a mixed session.

Peer beyond the numbers and it would have been readily evident that an 11-cent rise by big-cap SingTel to $2.76 pushed the STI up 11 points, thus creating the mistaken impression of a slightly firm or mixed market. Assuming SingTel had not risen and everything else remained the same, the STI would have recorded an 11-point fall, leading a casual observer to correctly surmise that the market had been weak that day.

Similarly, on Dec 29 last year, a sudden 87 per cent surge by CapitaMall Trust in the final minute of trading helped push the STI up 54 points, once again creating the mistaken impression of a session that was much stronger than it really was.

Still, using market-cap weights is probably a much better way to capture what's going in a stock market, at least for most of the time and over longer time periods. The alternative is to use price weights, which has been shown to lead to even more inaccurate conclusions.

On this last point, local investors - chartists and fundamentalists alike - would do well to take into account just how distorted a picture the price-weighted DJIA paints of the US economy and market, while also pondering whether 8,000 is really where its 'support' lies. If Dow at 8,000 is artificial, where does this leave the STI?