Anatomy of an event shock

Posted by Scriptaty | 8:58 PM

It shows the September 2004 euro futures (ECU04) contract. This chart shows the price action in terms of the euro, so “up” on the chart means the euro price is rising vs. the dollar.

At the time of the payroll shock the market was giving due consideration to the possibility of the dollar starting an up move. The dollar started moving up (and the euro down) in mid-July.

Remember, the euro down move occurred in the context of a longer-term up move corralled in a wide trading range that was congested and choppy at the same time.

Two days ahead of payroll and four days ahead of an expected Fed rate hike — both supposedly regular news stories — the euro failed to match a previous low (horizontal line). It came within a whisker, but was still 23 points (more than 20 percent) away from the previous low in a technical environment where the average high-low range was 109 points. The inability to match an old low reveals the dollar-negative bias was still alive and kicking.

If the non-farm payroll number had been within spitting distance of the consensus, those 23 points might have vanished and the down-sloping standard error channel might have been vindicated. In fact, the euro uptrend would probably have reversed to a downtrend once the old low was taken out, because a whole slew of indicators would have turned negative, too. This would have shown the upcoming rate hike was the decisive factor, after all, in its own right and also as a symbol of robust growth. You can think of the potential reversal of the negative bias as requiring two favorable factors, one cyclical and one institutional.

But the day before the big news, the market made a one day bet against the dollar. (It might have been a bet the payrolls news would be bad.)

Price closed above the open, breaking the standard error channel and carrying the relative strength index (RSI) close to its midpoint.

At this point, however, a robust payrolls report could still have made the euro bulls who bought on the close regret their bet. In fact, the day of the announcement, the price opened lower than the previous close (and traded lower) before turning to the upside.

In short, you could not have known from the price action in the few days leading up to the employment report (and even on the day of its release) which way the price might break on the news. This is the essence of a consolidation, which often reflects indecisiveness. There is still a preponderance of lower euro lows, which is an important characteristic of a downtrend, but it is offset by the failure to match or exceed the previous low, which suggests the down move is over. So, what action should you take?

Bottom line: There’s no technical or fundamental basis for a decision until the news actually comes out. It could go either way. Price is not providing a clear reading of market sentiment. A good payroll number would reinforce the case for the rate hike, and together these factors might feed the dollar reversal scenario. A bad payrolls number might not stay the hand of the Fed, but it kills the dollar up move.

And that’s what happened. The payroll number turned out to be so bad it became an event — as evidenced by the abnormally wide-range bar that broke the top of the average true range channel and carried the RSI above 50. A wide-range bar is one that is a certain percentage (e.g., 50 or 100) larger than the average bar over the previous n number of days (e.g., 15). In this case, the bar was 200 points when the recent 15-day average was 109 points. A widerange bar is hardly ever a one-time anomaly when it is associated with an event.

The expected rate increase from the Fed materialized and the euro put in a close at the low for one day — a normal, news-triggered outcome, but not an event. A few days later (and one week after the employment report) came the coup de grace — the trade deficit hit a record high, reminding everyone why they had acquired a dollar-negative bias in the first place three years ago.

Trade numbers have a curious track record in forex analysis. They are a key variable in fundamental cyclical analysis, and yet they hardly ever produce an “announcement effect.” This time, though, the reaction was tremendous and constituted a second event. The euro hit a nearly one-month high and closed at its high for the day. Now all doubt was removed. The dollar bears/euro bulls won. The new consensus was the bias is back — big time — and the euro would target new highs. The wide-range bar effect is somewhat reliable, but you absolutely must know the bias embedded in market sentiment ahead of time and how the cyclical data fits into it. Of the 39 wide range down bars over the past four years, the euro was trading lower 15 days later only 40 percent of the time. Of the 57 wide-range bar up moves over the past four years, the euro was higher 15 days after the big-bar day 60 percent of the time.

Here’s where things get a little tricky. If you know a news announcement will be an event shock if the data deviates wildly from the consensus estimate, you can forecast a wide-range day and, thus, take the trade in the direction of the bias on the same day as the announcement with a better than 50-50 chance of being right, not only on the same day but also 15 days later. (And traders know this, which is why profit taking in the euro after the wide-range days on the chart was so modest.)