Monday, 16 April 2018


I'll start with this handy little diagram:

The 'Rounds' indicate distinct trading phases.

In hindsight, didn't it all look so simple? So obvious? Well, no.

Hindsight bias is inherently a dangerous thing. It reinforces your inflated sense of self belief and prompts you to cherry pick the facts. You may even ignore serious flaws in your thinking process indefinitely. For the record, I did not foresee that such a big shift in GBGAQRS-WA was imminent (I anticipated a much smaller, orderly move) ; in fact for about 80% of the time I was only thinking of the downside.

In trading, making mistakes every day is how you learn. Losing money is how you learn; there's really no way around that.

1) You can be right about your analysis but you're way too early; hence the losses.

2) You can be wrong about your analysis and still make money out of sheer dumb luck.

3) You can be right about your analysis but your trading execution is poor; hence you lose out on the majority of the profitable price shift.

4) Your execution can be at the level of algorithm-driven high frequency trading computers, but your idea sucks; hence you lose a lot of money.

Which one would you rather pick? I'll tell you. It's (1) and (3). Execution is something you can work on slowly and steadily. Having the wrong ideas at the start dooms you to losses. If you like the thought of (2), well, there are casinos for that.

This post is mainly about (3), the only option where you will not lose. Obviously I made some good money by virtue of being early during this move, but that thought process has been expounded in my previous post on GBGAQRS. This time I managed to miss out on a large part of the move due to poor trading execution.

Here is where I will bare all my successes, mistakes and what can be learned from them.


For the record, I'm extremely enthusiastic about the use of technical analysis. Using past price activity to predict future movements is a waste of time for me - but I may also be the worst technical analyst in the world.

 This. From Scott Adams.

What I excel at is analyzing current price and volume activity. This is done by repeated observations and finding market signals; there is really no other choice to do this properly without sacrificing the time and effort.

My niche is in warrants - I'm fairly adept at finding the best warrants to trade (the one with the best profit potential on an absolute return and percentage return basis; here's a guide and checklist) and analyzing the discrepancy between the warrant and the mother share (which was the biggest thematic angle in GBGAQRS-WA).

But I do not completely discount the value of technical analysis. At best, it tells me what others in the market are thinking. Support and resistance points are no theoretical concepts; they are tangible and many traders swear by it.

To me, they are just helpful guides in my trade positioning. I can find price points where I can anticipate a large buy or sell volume; this is very helpful when I'm unloading or accumulating a substantial position (I currently trade in the hundreds of thousands of lots for a single idea).

And with that, let's look at a bunch of charts. The following is how I traded GBGAQRS-WA and how you can trade it better than I did.

Huge dip on 4 April; it turned out to be a great buying opportunity.

I'll spare the technical analysis mumbo jumbo; the last big candle (on 9 April) in this move basically indicates a breakout. It would have been wise to maintain the position, but I sold my entire position at the 35.5 sen mark.

Just going back to the first diagram at the beginning; this is how my trades unfolded in three distinct rounds:

ROUND 1 : 30 March to 2 April (two days). Bought at 23.5 sen, sold at 32.5 sen (x00,000 lots). 38% gain.

--------------  5 April : My first blog post on GBGAQRS.

ROUND 2 : 6 April to 9 April (two days). Bought at the 27.5 sen to 30 sen range. Sold at 35.5 sen (x00,000 lots). 24% gain.

ROUND 3 : 11 April to 12 April (two days). Bought at 39.5 and 40 sen. Sold at 44.5 and 45 sen (100,000 lots). 12.5% gain.

Here are a few useful observations on my trading activity over this two-week period.

1) ROUND 1 and avoiding the 4 April rout : this move demonstrates the value of short term trading. It was a twitchy market during that week and I was a twitchy trader; I was eager to get out of this trade ASAP. I opted for this instead of building a massive position and holding on to it. During the rout on 4 April, when a massive late afternoon selloff sparked a 2% decline in the KLCI, I was already out.

When trading big lots, you have to factor in two things : opportunity cost and liquidity risk. My gains would have been totally wiped out had I not exited the position on 2 April. And because of the sheer volatility on 4 April, there was no telling whether the stock and warrant would completely collapse on 5 April.

Round 1.

It also demonstrates the value of having a concrete plan and sticking to it. I typically aim for a (lofty) 30% return on investment for a single trade, but having a leveraged position means my profit target is halved at 15% (all things considered, the absolute value of the capital return is essentially unchanged due to a larger position thanks to leverage). In this particular scenario I was still able to exit at a 38% gain, far beyond my expectations.

2) PRICE AND VOLUME TELLS YOU ALL YOU NEED TO KNOW (5 April - 6 April). It became apparent quickly that the 4 April rout was completely overblown. The next day GBGAQRS-WA (which fell 25% on 4 April) quickly stabilized; there was some concrete buying interest and accumulation activity.

The following overlap between GBGAQRS stock and GBGAQRS-WA is a simple indicator that the warrant's valuation is overstretched.

On that day the mother share closed at RM1.62. The warrant however closed at 24 sen. A discount in the spread suddenly opened up; theoretically the warrant should be worth 32 sen (RM1.62 minus the warrant's exercise price of RM1.30).

Obviously this correlation is imperfect - the warrant is never tightly correlated with the mother share in this way - but it became clear quickly that the spread has widened. Prior to this, in recent weeks GBGAQRS-WA was trading at a 3-4 sen discount to what it should theoretically be relative to the mother share. When this discount suddenly widened to 8 sen, a real trading opportunity emerges.

So, let's think in terms of scenarios. Let's say I initiated a position at 24 sen per warrant:

SCENARIO A : The 8 sen spread returns to its historical average of 4 sen. This means I can easily derive a 16% gain from just the narrowing in the spread.

SCENARIO B : The mother share could suddenly break the tight RM1.65 range where it has been trading over the past few days. This means a 16% gain is merely the tip of the iceberg. If the stock moves to RM1.70, assuming a 4 sen discount in the warrant, GBGAQRS would be worth 36 sen. So the profit range is now between 16% and 50%; astounding numbers.

 GBGAQRS shares finding some stability.

SCENARIO C : And then there's the holy grail scenario: a sudden rally in GBGAQRS beyond the RM1.70 mark could  spark an immediate reaction in the warrant. If this happens, there's a real possibility that the warrant could rise and reach 'parity' - in this case at RM1.70 the warrant could be fully valued at 40 sen, thus eliminating the historical discount.

3) ROUND 2: I had enough capital in my trading account to take possession of the warrants (instead of thinking in contra trading terms). I was also impatient to take profits, but this was a reasonable move given that my position was leveraged. The problem with trading huge lots in warrants like these is that liquidity comes and goes. It's mainly present during big rallies, not during a selloff. It is fairly difficult to unload 300,000 warrants into the market given GBGAQRS-WA's typical liquidity.

What was the consequence of not staying in? I missed out on a big move from 36 sen to a peak of 55 sen in two days. That's a 52% gain. And based on my previous position, size I missed out on about RM25,000 in potential profits. This pretty much sums up good ideas but poor execution; don't feel sorry for me though.

But beware of hindsight bias; it did not look clear at the time that such a big rally was imminent. I had a plan and I stuck to it; for ROUND 2 my gains amounted to about 24%, or more than adequate for a couple days' work. The market was still volatile and all I had in mind was capital preservation. Profits typically take care of themselves but losses never do.

Round 2.

4) ROUND 3 : By 11 April, it was clear that the best case scenario has happened - SCENARIO C. Not only was there a huge influx of buying interest by now, GBGAQRS-WA was recording crazy one-day gains (33% on 10 April, 18% on 11 April). I missed out on all this of course.

 The best case scenario totally happened.

But if there was a clear-cut sign that the rally is slowing down, it's this:

a) After three days of consecutive gains there will be a great desire to take profits. Those who entered at the 50 sen mark were clearly late in the game.

b) 'Parity' is reached, contrary to my expectations. At GBGAQRS-WA's price peak of 55 sen on 11 April, it was trading at parity to the mother share (RM1.85, or RM1.30 plus 55 sen). When this happens, it's time to retreat; the warrant has achieved its full potential.

c) The warrant's percentage gains has steadily outpaced the mother share's. This indicates greater enthusiasm in the warrant, but at a high price point, there's nowhere to go but down.

So guess what happened on the afternoon of 11 April? The warrant lost its liquidity, leaving many latecomers trapped. In the morning session there were tons of enthusiastic buyers of the warrant at the 50 sen and above mark. By 4:30PM it collapsed to 40 sen, or an immediate 20% loss.

Having been a trader for a few years, and having been a part of the panic-selling crowd in the past, I can sense selling desperation when I see it. This was the first component.

The second component is probability assessment. There's little likelihood that the mother share would stage a similar 20% price collapse (its prices are easy to maintain, and the stock tends to trade in a tight range). When GBGAQRS-WA hits 40 sen, the prevailing assumption is that the mother share would fall to RM1.70 (assuming parity is maintained).

The stock did fall to an intraday low of RM1.74, but even at this price the warrant would fetch a valuation of 44 sen. Instead, there was some clear panic selling at the lower range of 39.5-40 sen.

I scooped up 100,000 warrants at that price. At the end of that day, the mother share closed at RM1.77, while the warrant closed at 43 sen. If you didn't notice, I'll tell you; the historical 4 sen discount has returned. The market sees fit to value the warrant at this level again.

On 12 April I sold my positions quietly at 44.5-45 sen for a quick 11% gain.

Always sell against a rallying market.


In my first post, I wrote that the company can be considered a good proxy for the upcoming general election. They still are, and you can reasonably hold a long term position in the stock.

While in recent weeks there has been a general rebound in sentiment towards stocks, as we get closer to the election date things might be less clear from a risk-reward standpoint. If the KLCI keeps rising prior to the election, the post-election upside potential would be far less. In fact, there would be serious downside potential, especially if the GE results swing another way (I'd never rule this out).

This trade was the kind of situation where you can miss out on huge rallies and still make good returns. My total winnings from this?

My point is this : the easy money has been made.