The Way of the Dollar

fluctuation. Commercial or investment

demand may drive price action in the middle stages of a price movement; and this kind of demand is not necessarily price-led. But as trends gather momentum, speculation becomes the dominant driving force behind price and causes the extreme. A price chart can tell us that a price has had an extended move, and that’s all it can tell us; whereas the open interest can indicate to us whether the move is over-extended. It can tell us whether a currency is overbought*, or oversold*.

In the terrible UK bear market of 1974, Burmah Oil was forced to dispose ofits giant stake in BP, since it had borrowed hundreds of millions of pounds on the security of its BP share-holding, and the slide in the price of BP was literally threatening Burmah with bankruptcy .The day after Burmah sold its holding early in 1985, the whole stock market turned on a dime; and within a few weeks the price of BP had doubled again. This wasn’t chance. The threat of Burmah’s forced sale was itself responsible for the last burst of pessimism and bearish speculation in that long bear market. Hence news of the sale turned the tide.

Remember the last time you sold a currency at what proved to be the bottom, or bought at the exact top? That wasn’t just bad luck – nor even just foolishness. You and the crowd caused the bottom, or the top.

The Consensus*

I know of one top equity fund manager who has no other rule for handling the currency markets than to go against the consensus. It’s common sense. We must be ‘contrarians*, if we are to survive in financial markets in general and the currency markets in particular. During the great bull market in the dollar in 1981-5, it was the one single rule that assured survival. If you have any problem with that, I suspect there is no solution but to observe markets till it’s no longer a problem; for the shifts from pessimism to optimism and back are what bull and bear markets are about.

The difficulty is to define “the consensus*”. The crowd isn’t always wrong. When a price movement gets going, the crowd as often as not will line up in the direction of the movement. But standing against the crowd, at times can be as desirable as standing in the way of an express train. This is the drawback of such objective measures of opinion as Market Vane* – a well known American service which measures bullish opinion among traders. If you poll traders, most of them will point in the direction of the trend. Bullish opinion as measured by Market Vane tells us the direction of prices over the past week, but not necessarily a lot more.

Perhaps Bruce Kovner*, of Caxton, nailed the problem when he said (see Chapter 8) that what he was looking for was the consensus that is not

confirmed by price action. That covered the entire 1981-5 bull market in the dollar when the consensus was constantly bearish. It also covers the price extremes* , when the consensus is wrong by definition.

Whether we are looking at the underlying multi-month/multi-year trend, or the intermediate multi-week moves, there are usually two phases when the consensus is not confirmed by price –early in the move and at the end of the move. Soon after a price reversal, majority opinion is usually aligned with the previous trend, i.e. the consensus lags. Similarly, majority opinion strengthens along with the on-going trend, tending to reach peak consensus at the price extreme. So the ideal position is to be contrarian at the beginning and end of a move, and pro-consensus in the middle. Nice work if you can get it so right.

Never forget that the consensus usually includes you.

The consensus gauge is a subjective gauge. We read the papers and specialist commentators and we talk to people, and we conclude that most punters are facing one way. If we are facing the same way, we have to reconsider the situation in the light of our other sentiment gauges and cut back if they are flashing yellow. In the heat of a powerful favourable price move, we are often lulled into complacency: at that point, consulting the consensus is an essential discipline – it often comes as a shock to discover that we are in with the herd, and it can be very costly if we fail to make this discovery. When we diagnose a situation where the consensus is not confirmed by price, we should not just cutback but try facing the other way, to see whether anything clicks. If the market action* feels right; if the open interest is extended; and if we can find a fitting rationale, we can reverse our position.

Perception of the trend

The disadvantage of the fractal nature of market fluctuations is that you can get in a muddle over the underlying trend and its minor and medium corrections* and extensions. One solution is to simplify and reduce the problem to just two constituents –the major underlying trend and its main corrections, which are the medium-term, multi-week/multi-month corrections. That way all we are concerned with is the main trend and corrections to the trend.

The art of analysing financial markets is always to be able to reduce the problem to a simple black and white, yes or no issue. That way, if our analysis tends to be right, what we are doing is shifting 50/50 probabilities

into 55/45 or 60/40 probabilities in our favour. And if we can always do that, we must win in the long term. (The temptation is to make complex bets that X will happen and that y will happen: those are 25/75-type bets, of which we should have no part). What happens when a market trends is that more and more participants are being converted to the view that the trend has changed. Our perception of the trend gauge is designed to measure this conversion flow*. We have no statistics to go on –just feel. We can, of course, see the past direction of prices in charts, and we would be foolish not to use charts for all the help they can give us in gauging the perception of the trend. For a start, they can help us gauge the perception of chartists, who are significant players. But what we’re interested in is the changes in perception; and we try to feel these from anecdotal evidence before they are apparent in price histories – using price histories for corroboration.

When a trend is ready to change – from up to down, say – everyone is ‘facing’ up: everyone is projecting higher prices. Normally it takes time for the perception to change. Instead of higher prices you get a “double top”, or a couple of descending highs and lows. The perception of the trend does not change at the extreme of the old trend, but progressively after it. We have been alerted to the extreme by our open interest gauge, with luck. Then our monitoring of trader sentiment tells us that a shift is taking place: the consensus is being converted from bullish* to bearish.

Reaction to news

An early sign of a change in the perception of the trend is the way the market reacts to news. During the bullish phase, prices tended to rise on good and bad news alike. Then there is a change and good news fails to help the market. Finally, prices fall on good and bad news alike, and the trend has well and truly changed from up to down. The same applies in mirror image in bear phases.

Let me say that I am proposing nothing original here. This sentiment gauge is second nature to all old hands in financial markets. When traders say “the market is acting badly”, this is what they mean: they are not as