3 years ago, I embarked on a new strategy of placing small, speculative bets into loss-making companies with the potential to make a turnaround. That strategy is now affectionately known as the "minion" strategy. The first class of minions is from the semiconductor sector, which has risen strongly this year. Most of the minions have been sold in the last 1 year, and they have graduated with flying colours.

The triggering point for initiating this strategy is that I realised that although there were many semiconductor stocks listed on SGX, only 2 made good profits and gave out good dividends. The vast majority were not. See the table below, which is based on the financial results for FY2013, which were the latest available results at the time when I initiated the strategy in Mar 2014. Please note that the figures are not adjusted for consolidations and other corporate actions.

The intriguing question I had was why both Micro-Mech and UMS could make money but the rest could not. Some even had fairly large losses. The divergence in performance raised an interesting question, which was that would Micro-Mech and UMS follow the rest into losses, or the rest would follow Micro-Mech and UMS into gains. Thus, I decided to explore placing speculative bets into the loss-making companies, with the hope of them turning around and becoming multi-baggers. These bets were mentally written off the moment they were invested (see Meet The Minions for more info).

Having said that, it is not just anyhow throwing money away. Nobody likes to really lose money. Thus, there are 2 guiding principles in the minion strategy. Firstly, the companies must demonstrate they have the ability to survive at least for the next few years, so that there is sufficient time for a potential turnaround to happen. They should also not have to call a rights issue, else it would be throwing good money after bad ones. Secondly, there must be reasonable probability of a turnaround happening. If either of these 2 conditions are not present, the minions would likely lead to losses.

On the ability to survival, the companies should not have high Debt/Equity ratios and the Net Asset Value should be sufficient to absorb the loss per share for the next few years. Surprisingly, all the semiconductor companies evaluated above had low Debt/Equity ratios, with the exception of STATS ChipPAC. The NTA/EPS ratio for loss-making companies show how many years they could last, assuming they continue to make the same losses every year. Again, in this aspect, all the companies could survive for the next 4 years at least.

On the probability of a turnaround, I really had no insights into this industry (and why Micro-Mech and UMS made money but the rest did not) and was relying heavily on the guess that convergence among the companies (in either direction) was probable. Exactly how long the turnaround would happen was unknown. Based on the earlier discussion, if the turnaround were to happen within the next 4 years, then all the stocks evaluated would rise.

Besides checking whether the companies could survive, I also considered if a turnaround were to happen, how much money could the companies make. This is where the highest EPS in the past 5 years came in. It is not much use if the companies only made small profits at the peak of an industry cycle.

Next, the stocks must be selling at a cheap price relative to valuation. If they are not cheap enough, the profit potential is reduced. This is why even though Micro-Mech and UMS are profitable companies, they do not make good candidates as minions. The Price/NTA ratio shows that most of the companies have low P/NTA ratios of around 0.50, except for the 2 darlings which are Micro-Mech and UMS. In particular, Sunright only had P/NTA ratio of only 0.20.

Finally, diversification is extremely important. Despite all the checks, I cannot tell for sure which stocks would tank or call a rights issue. To manage this risk, I buy more than 1 stock.

Based on the considerations above, I selected ASTI, Ellipsiz, MIT (Manufacturing Integration Technology), STATS and Sunright for my speculative bets in Mar 2014. Each position was a small one, and I had 5 stocks to spread out the risks. Most of these stocks were sold in the last 1 year. The results are as shown below.

Company	Bought	Sold	% Profit	Remarks
ASTI	$0.055	$0.056	2%	Sold in Apr 17
Ellipsiz	$0.283	$0.380	34%	Sold in Sep 16
MIT	$0.066	$0.220	233%	Partially sold in Jul 15
STATS	$0.335	$0.625	87%	Sold in Sep 14
Sunright	$0.125	$0.305	144%	Sold in Mar 17
Average			100%

Among the 5 minions, 1 is a dud, 1 is a 2-bagger and 1 potentially could be a 3-bagger (assuming fully sold at the current price). The average gain is 100%. Many of them rose further after I sold.

So, the above are my first class of minions. They have graduated with flying colours and gave me enough confidence to continue my minion strategy.

Just a final note, in case you go away thinking minions are very profitable, they are actually high risk, high gain positions. Not all will make money. Some will show unrealised losses for long periods of time. Some will be completely wiped out. One of them, Ezion warrants, got suspended the day I bought into it. Do not attempt this unless you fully understand and are prepared to take all the risks.


See related blog posts:

• Meet The Minions
• The Minions (Millions) Mentality
• How to Get a Multi-Bagger?