Skirts up, taxi chat, buy, sell: here's the running inside out
They include the hemline index: skirts up, market up. Who wins the Super Bowl: if a National Football Conference team (NFC) wins then the market goes up that year; if an American Football Conference (AFC) team wins, it goes down. This indicator has an 83 per cent correlation apparently.
Then there are restaurant prices in New York. The higher they are the more dangerous the market. Then there are CNBC ratings. When CNBC ratings exceed soap operas the market is about to peak.
Then there's an indicator you can try at home. It's the "Cocktail Party Chatter Index". It works the same for the sharemarket as it does for the housing market. If your guests arrive and leave and never mention either, buy both. If they walk in going, "How's your portfolio going? Lovely house, what did this set you back?", sell.
Most of these indicators can be dismissed, of course, under the "sun spot" principle. In other words they rely on the principle that when X happens Y happens so X must cause Y. It is one of my bugbears, along with people saying "Bless you" or whistling inanely while they do a menial task, people in the stockmarket with too much time on their hands purporting to add value by making stockmarket predictions out of past statistical coincidence.
Of course, the beauty of these indicators, and why they endure, is that they are simple. Humans, faced with uncertainty and questions, constantly grasp at the simple solution. We really want to believe that small effortless steps like getting our credit cards out will have a deep and lasting impact. I myself have a Fitness First membership. Haven't been in over a year.
Here are some more ordinary indicators that have developed to guide us in the stockmarket. Some useful, some humbug. Your choice.
■January effect: The statistically correct phenomenon that share prices go up in January (and December). September, by the way, is the worst average month.
■Sell in May and go away and come back on St Leger day: I reckon the only reason this stockmarket idiom survived was because nothing rhymes with April. Despite that, the sell in May maxim is, unbelievably, statistically correct, but it relates not to the market performance in May but the market performance in the six months between May to October compared with the better six months from November to April. St Leger day is a group 1 horse race in the UK every September that marked the end of the British social season and time for the aristocracy to return to more financial matters.
■Mean reversion: The idea that stocks that go up a lot will then fall, that stock prices revert to the mean or average. This is statistically wrong in the stockmarket. In fact, rising prices correlate more highly with future price rises than future price falls. In other words, you should buy what's going up and sell what's going down because what goes up doesn't come down, it goes up.
■P/E and yields: PEs and yields only tell you what analysts are currently forecasting, but the real money is made in the stocks the analysts have got most wrong. PEs and yields don't help you with that. Buy cheap stocks and you will be buying all the stocks people don't like. They are cheap for a reason, usually because the earnings forecasts are too high.
■Volatility: Low volatility is an indicator of good future returns, high volatility is an indicator of poor future returns. Statistically a sharp rise in volatility can be associated with a sharp drop in prices and a fall in volatility correlates with subsequent rises in prices.
■Insider buying: Aggregate buying of directors of their own companies is seen as a lead indicator of equity markets. The "insiders" know before we do. Makes sense. In fact, talking to an experienced resources director the other day this was one of his abiding trading principles for resources stocks. Sell when they do.
■Taxi drivers: If they so much as mention the stockmarket, sell!
■Newsletters: Another contrarian indicator. The more bullish they are the more negative the outlook. I'm a bear at the moment, which is about as bullish as it gets.
All that aside, there is still one fail-safe indicator, proven to be 100 per cent effective in 100 per cent of cases. Hindsight. Harry is never wrong.
Frequently Asked Questions about this Article…
Colourful sharemarket indicators are simple, often quirky signals people use to predict markets — things like the hemline index, who wins the Super Bowl, restaurant prices or even CNBC ratings. They endure because humans prefer simple rules when facing uncertainty, but many of these indicators rely on spurious correlations (the article calls this the “sun spot” principle). Everyday investors should treat them as entertaining guides at best, not as reliable investment rules.
The hemline index is a tongue-in-cheek indicator that associates shorter skirts with rising markets (“skirts up, market up”). It’s an example of a simple cultural correlation rather than a causal economic factor. The article presents it as fun but not a robust basis for investment decisions.
The Super Bowl indicator links the NFL champion to market direction: if a National Football Conference (NFC) team wins, the market supposedly goes up that year; if an American Football Conference (AFC) team wins, the market supposedly goes down. The article notes this indicator has been claimed to show an 83% correlation — a striking statistic, but still an example of historical coincidence rather than proof of causation.
The Cocktail Party Chatter Index is an at-home contrarian indicator: if guests never mention housing or portfolios, buy; if everyone walks in talking about their portfolios and how much they paid for houses, sell. It’s a light-hearted way to spot overheated public sentiment — useful as a contrarian red flag, not a strict rule.
Both sayings have a statistical basis but need context. The January effect notes a tendency for share prices to rise in December and January (with September historically a weaker month). “Sell in May and go away (and come back on St Leger day)” survives partly as rhyme, but is statistically tied to the relative performance of the six months May–October versus November–April — not necessarily to returns in May alone.
According to the article, mean reversion — the idea that big winners will fall back to the average — is statistically wrong in the stockmarket. Rising prices historically correlate more with further rises than with falls. The article suggests a momentum-style takeaway: buy what’s going up and sell what’s going down, rather than assuming winners must revert to the mean.
The article suggests P/E ratios and yields only reflect current analyst forecasts and don’t reveal where analysts are most wrong. While they provide context on valuation, the piece warns that buying very cheap stocks often means buying companies that are unloved for a reason (usually overly optimistic earnings expectations). In short, P/Es and yields are useful data points but not foolproof stock selectors.
The article highlights a few straightforward indicators with statistical backing: volatility (low volatility tends to be associated with better future returns; sharp rises in volatility can accompany price drops) and insider buying (directors buying their own company stock is seen as a lead indicator). It also notes contrarian cues such as overly bullish newsletters or everyday chatter (including taxi drivers talking markets) — these can signal crowded optimism. Finally, the piece reminds readers that hindsight is always perfectly accurate, so be cautious about treating past patterns as guarantees.
Call 1300 880 160
Start a chat
Schedule a call
