BG roller-coaster ride finally pays off, thanks to Fibonacci

Last week, things looked bad. I had managed to turn a profit of more than £2,000 (Dh11,584), carefully (or, more accurately, luckily) crafted over 25 weeks, into a near-loss.

Powered by automated translation

And that, ladies and gentlemen, is what we call vindication.

Last week, things looked bad. I had managed to turn a profit of more than £2,000 (Dh11,584), carefully (or, more accurately, luckily) crafted over 25 weeks, into a near-loss.

In a sense, I was in a trap, tied to the roller-coaster ride of a single stock - the very opposite of the bet-hedging diversification urged by every idiot's guide to day-, or swing-trading.

But, as I explained last week, I wasn't simply holding on to my shares in BG, the former British Gas, out of stubborn pride; I was attempting to apply some science to my blind faith and guesswork in the shape of Fibonacci retracement patterns.

This mysterious maths-based analysis of the impact of human behaviour on the price of shares - the way in which herd instinct, at predictable moments in the price cycle of any stock, will coagulate into a buying and selling pattern for no discernible, exterior reason, and effect a price movement - appeared to indicate that BG's price was about to change direction. And, lo, it has come to pass.

And not before time. Last week, I - or, rather, Fibonacci - predicted that, despite the apparent downward spiral in value, BG would soon pick up. I didn't really believe it, to be honest, but now my faith in 13th-century Italian mathematicians has been restored.

I'm sorry to gloat, but here I'm going to repeat what I wrote last week: "If human behaviour is all it's cracked up to be, then we should next expect a sell-off - which will in turn provoke a buying spree, raising the price once again."

And what have we seen? That's right, ladies and gentlemen, exactly that, before your very eyes.

Overall, in a series of purchases spread over a few days, I had paid an average of 1,242.7 pence for my BG shares, the price of which last week had dropped to a mere 1,168.5p. Applying Fibonacci's numbers theory - applicable to everything from rabbit-breeding patterns to stock-market share movements - I boldly predicted they would soon bounce back to about 1,200p.

I was wrong, but in a very, very good way. In fact, they bounced back to 1,267p.

Now, of course, was the time to sell, to make hay while the sun was shining - but to do so intelligently. A new trading day loomed. There would be nothing worse than to miss the moment and see the price drop rapidly back down again at the opening of business, so to protect against this, I put a stop-loss in place, to sell the lot if the price dropped to 1,260p.

Then I had a closer look at the chart for the past three months - and I've reproduced it below so you can see what's been going on. The red bars show selling activity; the green, buying. Selling provokes a drop in price, buying, the opposite.

Take a look at the buying activity that has stimulated the current price restoration. Now look back along the graph; it's the most sustained buying spree in the past three months. That can mean one of two things; either the buying is nearly over, and we can expect a brief period of consolidation followed by a fall, or this is the start of a major upwards price adjustment.

I decided to put in place stop-loss and limit positions that would protect me against losing out in either eventuality. Of course, if the price really rocketed upwards, I was ready to adjust my limits accordingly.

But what should my upper limit be? If we really are seeing a major price recovery, how high could it go?

In the past year, this share hit its peak to date of 1,296p at the beginning of November. Step back further and look at a chart for the past two years, however, and it becomes apparent that this most recent peak was the third in a series that began in May 2009. Since then, the price has fallen away, then risen higher, fallen away again, and then risen higher still.

Drawing a line through each of these successively higher peaks leads to the tempting inference that, if we really have now embarked on the final ascent to a fourth peak, we will shortly be requiring oxygen, as the price could be on its way to 1,400p or more.

That price, however, seemed far too high, so I opted for a conservative upper sell limit of 1,300p.

In the event, I acted just in time, and it worked beautifully. The price flicked upwards and briefly hit 1,300p a share within a few minutes of the market opening. The software dutifully cashed in my chips, leaving me with £16,720.36 in the bank, representing a profit on my original £15,000 fantasy investment of £1,720.36. Not a bad recovery from last week's flat line.

If this were for real, it would have been a pulse-quickening couple of weeks - but I doubt my sense of self-satisfied smugness would be any more pronounced than it is now.