I think the price history graphs of speculative investments such as housing and gold look like a series of linked "u"s, ie sharp peaks and rounded troughs, while the graphs of less speculative investments that pay a dividend look like a series of linked "n"s, ie sharp troughs and rounded peaks. Do you agree? Has this been observed before? I have my own ideas, but why do you think this might be?

dow jones

gold price

uk house prices corrected for inflation


Let me see if I can restate your question: are speculative investments more volatile (subject to greater spikes and drops in pricing) than are more long-term investments which are defined by the predictability of their dividend returns?

The short answer is: yes. However, where it gets complicated is in deciding whether something is a speculative investment.

Take your example of housing. People who buy a house as an investment either choose to rent it out (so receive "rent" as "dividend") or live in it (foregoing dividends). Either way, the scale of the investment is large and this is often the only direct investment that people manage themselves. For this reason houses are bound up in the sentimental value people attach to a home, the difficulty of uprooting and moving elsewhere in search of cheaper housing or better employment, or the sunk cost of debt that can't be recovered by a fire-sale. Such inertia can lead to sudden sell-offs as critical inflection points are reached (such as hoped-for economic improvements fail to materialise and cash needs become critical).

At different levels that is true of just about every investment.

Driving price-volatility is the ease of sale and the trade-offs involved. A share that offers regular and dependable dividends, even if its absolute value falls, is going to be hung on to more frequently than those shares that suffer a similar decline but only offer a capital gain. For the latter, the race is on to sell before the drop neutralises any remaining capital gain the investor may have experienced.

A house with a good tenant or a share with stable dividends will be kept in preference for the quick cash-return of selling an asset that offers no such ongoing returns.

This would result, visually, in more eratic curves for "speculative" shares while more stable shares are characterised by periods of stability interspersed with moments of mania.

But I have to take your query further, since you provide graphical evidence to support your thesis. Your charts combine varying time-scales, different sample rates and different scales (one of which is even a log scale). It becomes impossible to draw any sort of meaningful micro-comparison unless they're all presented using exactly the same criteria.

  • 1
    My question is not about the degree volatity. It is about the shape of the volatity. And yes, I agree that the evidence is not well presented: If I have the time I'll try to do something about that. – Michael Burrows Sep 20 '10 at 14:46
  • If it's specifically about the shape then the alignment needs to be pretty good between the charts. That also includes having similar degrees of liquidity. A share can be sold almost instantaneously. A house ... not so much. – Turukawa Sep 20 '10 at 15:09

I agree with @Turukawa that the x-axes need to be the same to make a direct comparison. However, the graphs you linked make me think of introductory calculus: If you time averaged plots, speculative investments (gold, housing) seem to have many large concave up time periods and the dow jones has many concave down sections. Using the concavity test:

  • Second derivative > 0 for speculative
  • Second derivative < 0 for less speculative

If the first derivative tells you about the rate of change, the second derivative tells you about the rate of change of rate of change. Remember back to Physics 101: 1st derivative is velocity & second derivative is acceleration.

It would be interesting to have the same time scales for your plots & compare these accelerations between the two. I suspect the more volatile investments would have larger (in magnitude) accelerations during boom/bust cycles than less speculative investments.

  • 1
    I like the concavity notion. I think the dow has many concave up and down sections, and housing has many convex up and down sections. – Michael Burrows Sep 21 '10 at 6:09
  • First "convex down = concave = negative second derivative" and "convex up = convex = positive second derivative". Second, stock price curves are too erratic in general to have derivatives. Even in the ideal case of continuous time trading, the models used are stochastic and produce at best continuous curves. Differentiable curves in stock prices must be a rarity. – Raskolnikov Dec 25 '10 at 13:42

Dividend-paying securities generally have predictable cash flows. A telecom, electric or gas utility is a great example. They collect a fairly predictable amount of money and sells goods at a fairly predictable or even regulated markup. It is easy for these companies to pay a consistent dividend since the business is "sticky" and insulated by cyclical factors.

More cyclic investments like the Dow Jones Industrial Average, Gold, etc are more exposed to the crests and troughs of the economy. They swing with the economy, although not always on the same cycle. The DJIA is a basket of 30 large industrial stocks. Gold is a commodity that spikes when people are faced with uncertainty.

The "Alpha" and "Beta" of a stock will give you some idea of the general behavior of a stock against the entire market, when the market is trending up and down respectively.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.