Lessons Learnt: Hyflux - Part 2

This is a continuation of the previous post.

In the last post, we summarized the Hyflux debacle and discussed the lessons learnt.

1. Always limit or risk capital to an amount we can lose

2. Map out all the scenarios and probabilities and keep monitoring them.

Today we delve in #2 and share the third and fourth lessons further below. In our due diligence on Hyflux eight years ago. I would say that I did not do this well. Hyflux scenarios and probabilities should have looked like the following:

60% - business as usual, Hyflux continued to operate as successful as it had since IPO. In this scenario, things pan out as we wanted, perp holders get back their money when Hyflux redeemed them in 2018 or 2020. This would be the base case or good scenario.

30% - Hyflux business deteriorates or the external environment changed, causing some cashflow problems. But Hyflux should manage to pull through either with bank credit lines or with the Singapore government awarding it yet another project. Or Hyflux raised equity or debt to fund itself.

As a side note, this was partially played out in 2016 when it raised yet another round of perps. It should have served as a warning sign, but back then, no one suspected anything. Hyflux was going strong and Tuaspring was touted as a gamechanger. 

10% - Hyflux fails for some reason. This is the worst case or disaster scenario. In Hyflux's case, this scenario is playing out now. 

Investors complaining Hyflux's epic failure

In my analysis, I did not pay attention to this last scenario. I would never had ascribed a 10% probability of failure at that time. But maybe it should be a 1% or a 5% probability event. I should have considered it. Let's for learning sake mapped out how things would be like if I had ascribed such a scenario. Say, we ascribe a remote scenario that Hyflux would fail. The probability that Hyflux would fail is higher than say, DBS or UOB would fail. Then logically it means that 6% is not good enough. This was because DBS or UOB perps were at 5%. Between Hyflux at 6% and DBS at 5%, which investment is better? I would say DBS.

Inverting the thinking a bit, the question should also have been asked clearly: at 6% yield, we get back our capital after 16.7 years. Will Hyflux fail in the next 16.7 years? Hard to say but if it happened would this still be a good investment? The answers are clear on hindsight, now that things had happened. It would be hard to answer these back then, but still, these questions ought to be asked. It could meant a different decision: not to invest at all, or bid lower and put less capital at risk, or maybe buy DBS perps instead.

There was also a lot hype when the offering was launched. It was way oversubscribed and hence such risk hedging thoughts were thrown out of the window. So, on hindsight, there is a sub-lesson here (which again, we already know): the crowd is not always right, be fearful when others are greedy.

The other mistake was also the lack of monitoring. After I bought these, I was just happily receiving coupons and occasionally read some annual reports and followed the news. That's it. I didn't even know Hyflux issued more perps in 2016 until a few months later. Then when things really turned south, I still wasn't monitoring as hard. This brings us to the third lesson.

3. Act fast don't hope

As things deteriorated. I held hope that Hyflux could turn around. The perp prices fell from 80c to 50c to the dollar. I could have sold! That way, with the coupons clipped over the years, I would have lost a mere 10-20% of capital rather than 60%. This would prove to be a lesson that I had not learnt well. I had never cut loss well. I cut losses only to see stocks rebound 50% and fail to cut those that go down a lot more. It would likely take more years to hone this skill better.

Hope is a dangerous thing

This has to do with judgement. In Hyflux's case, we had determined that the business model was flawed. It needed to bid for projects and cost overrun could be very detrimental. When they issued another perp in 2016 and when the initial warnings came, I should have paid more attention. Judgement can only be honed over time and experience. It is also about understand the business model, the situation and all that is at stake. When the bond prices finally reacted and fell to 50c, it was supposed to be a big warning. Yet I failed to do more detailed due diligence. It wasn't my priority until everything blew up. Hence there's a last lesson here from Hyflux.

4. Focus on the best businesses

Investing is a full time job. If you want to make money, you have to devote the time and effort. But in today's world, where got time? We have our day jobs, family, kids, friends, social and community activities. It's just so difficult. This is so even when I am passionate about investing. Imagine someone who is not passionate but wants to invest because he or she thinks it's good passive income, easy money.

So in order to be able to invest and sleep well, we can only buy the best businesses because we don't have time to monitor any deterioration. Having said that, good businesses also get disrupted and we need to stay on top of things when we see these happening (Singtel, which I have, comes to mind). If we buy the best businesses, those with strongest economic moats which we had discussed before, generating strong free cashflows at reasonable valuations, then our capital would have more protection. 

In conclusion, here's the four lessons again from the Hyflux debacle:

1. Always limit or risk capital to an amount we can lose

2. Map out all the scenarios and probabilities and keep monitoring them.

3. Act fast don't hope

4. Focus on the best businesses

Hope this would help us avoid future debacles, huat ah!

Lessons Learnt: Hyflux - Part 1

Eight years ago, we discussed Hyflux perpetual bonds here, putting forward the investment thesis that 6% was good dividend/interest income and how Hyflux had a so-so business model but things should be okay because the Singapore government would support Hyflux as they had done so in the past. That turned out to be a huge mistake. Not only did the government not support Hyflux, she rubbed it in, pushed the proverbial dagger into Hyflux's belly, delivering the fatal blow.

Et Tu Temasek? (Ref: Et tu Brute)

How did things come to such a dire situation? 

As described in the post eight years ago, Hyflux's business model relied on winning water projects, which meant that they had no control over the bidding price and also, in subsequent years, their own future revenue. However, as with most Singapore co.s, we are good at managing costs, which allowed us to beat many others in the global game of winning EPC (Engineering, Procurement, Construction) contracts. This was how Keppel and Sembcorp became so good in oil rigs.

But in order to grow, companies in the EPC field have to bid for bigger and bigger projects. The cost management however gets more and more complex. Once every decade also, someone would definitely screw up and one or two badly designed contract put EPC firms at risk of bankruptcy. Alas, Hyflux was not spared.

Tuaspring, Hellspring

The irony for Hyflux was that the contract turned out to be one in its home country. This was infamous Tuaspring desalination project. Tuaspring became a bomb because of its large scale and complexity. The Achilles' heel in Tuaspring is actually not desalination but power generation. For reasons unclear to me now, someone thought it's a good idea to combine the two. Maybe because desalination requires a lot of power, so hey why not generate power, then sell some power plus water to PUB as well. This definitely developed as the logical train of thought from our admin officers in the civil service and Hyflux went along.

But selling power is not the same as selling water. Cost for selling power depends on fossil fuel, the most important being crude oil, which is notoriously volatile. Meanwhile, power prices in Singapore collapsed as a result of energy deregulation. So suddenly, Hyflux found itself caught in a situation where its power generation cost exceeded its revenue. With a billion dollar debt on its balance sheet, things quickly spiralled downhill. Our admin officers don't give chance these days, just like the new Certis Cisco summon officers.

Hyflux began to run out of options as credit dries up and in an "unthinkable" scenario for investors who put in money in 2001 during its IPO to yield chasers (like me) who bidded to buy its perps in 2011 and more yield chasers who bought its perps again in 2016, Hyflux declared it might go bankrupt. A white knight from Indonesia (Salim group) appeared, willing to put money to save the firm but not the perp bondholders. Alas, that hope is now also gone as PUB decided to push the dagger some more (figuratively impaling Hyflux now), terminating Tuaspring water purchase agreement. 

As things stand, it is likely that bondholders and shareholders will get nothing in the end, barring some kind of miracle. It's neither Oliver Lum's nor PUB's fault. This is just investing. There's always risk. Perils on top of perils. Caveat Emptor.

The saving grace for those of us who invested in 2011 though was that we got a few years of 6% coupon. Not great, as the total loss of capital was still north of 60%. But almost everything is now bridge under the water. So this post serves to help us learn the lessons and move on. Well, it's mostly re-emphasizing the importance of what we already know:

1. Always limit our risk capital to any single name to an amount that we would be okay it if goes to zero. This could be an absolute amount and it could be a percentage of total net worth. What's important is that we can sleep at night. We hear stories of people who put in $200,000 or $300,000 into Hyflux perps and that's a huge chunk of their retirement nest egg or net worth. It's just so sad. So don't make this mistake.

2. Map out the scenarios and probabilities well and keep monitoring them. When we did the due diligence in 2011, we determined that the business model was flawed, one project could kill them. But we also thought that the Singapore government should come and bail them out. After all, this was Singapore's poster child. Little did we expect it would be the opposite! The project was in Singapore and the government delivered the fatal blow! Nobody could have foreseen this. But, on hindsight, we should have put in a scenario that Hyflux would go bust and ascribed a probability. In the next post, we should delve into this!

Caveat Emptor: Let the Buyer Beware!

Hyflux Preference Shares

Like most yield hungry Sg investors, I have been googling around on Hyflux's preference shares, but sadly, the information online is still somewhat lacking. One of the best analysis out there is done by la papillion:

http://bullythebear.blogspot.com/2011/04/hyflux-preference-shares-part-1.html

The blogger has written a blockbuster trilogy as well, do read through his whole analysis, totally worthwhile! Considering that the deadline for subscription is like 3 days from now, I guess this will be the first and last post for me.

For those really dunno what's going on, Hyflux is the leading water company in water deprived Singapore. The mkt cap of the firm exceeds a billion dollars, which means it is not your small fry SGX listed co, this can be the next Keppel Corp one! Recently, the firm announced this plan to raise S$200mn via pref shares with perpetual yield of 6%, with a step up to 8% some years out.

A pref share is similar to a bond in the sense that it gives you an interest income (well in this case a more or less compulsory dividend). The difference being that the dividend is perpetual (at least theoretically). Also pref shareholders are junior vs debt/bonds but senior vs common equity. This means that if the firm goes bust, pref shareholders stand to get something only if all bondholders have taken their share - ie nothing much left lah. However, pref shareholders usually enjoy a higher dividend. In this case, a good 6%!

As for Hyflux, there are various issues involved and I will simply summarize the pros and cons briefly here:

Here's the good part first.

1. The dividends are cumulative - ie if they fail to pay one time, they must make up next round, ie you will more or less get your 6% over time.

2. The dividend must be paid as long as common shareholders are paid, bcos pref shareholders are ranked higher vs common shareholders. Looking at its common dividend, Hyflux's track record is ok, not super stellar but at least got dividend every year since a couple of years ago, albeit the yield is pathetic, like 2% kind.

3. Hyflux enjoys the support of the SG govt, while it is unlikely that the govt will step in to save the stakeholders (ie debt or shareholders) in the event Hyflux goes belly up, it is also unthinkable that Hyflux would actually go bankrupt in the next 3-5 yrs. The govt has in the past been supportive of Hyflux, awarding landmark mega contracts (like the first de-salination plant etc) to the firm every couple of years.

4. It is a well-runned company. Top management knows their stuff and has earned the respect of both investors and competitors. Just that it is in a bad business. More on this below.

Here are the cons:

1. Why is the firm doing such a detrimental financing? They can always borrow from the banks at a lower rate, but they chose pref shares, and have to pay 6%. It might have something to do with their credit rating, if they raise more debt, the rating will fall. It then also says something about their financing options. Last checked, the firm has S$1bn in debt and S$500mn in equity. It's definitely higher than most SG blue chips although not that bad if you compare it with global co.s. Maybe the uprising in Africa is really taking a toll on their businesses. Or maybe Olivia Lum thinks that she should share her success with all Singaporeans. Well that is my wishful thinking :)

2. Hyflux's business model is not robust. It relies on winning projects, which in turn is economic sensitive. Even worse, its growth relies on winning ever bigger projects, most recent ones were like US$1bn or more one. In bad times, like during Lehman, project pipeline dries up and cashflow goes into red.

Also, they have no say over the price, they bid for the projects and can only win with lower bids vs others. Inflation in the last few years is causing their input costs to go up. Now with Libya, I suspect they are not in a good shape although a recent SG project should help. This validates the point made above: the SG govt lends a hand during bad times.

3. The cashflow track record is not exactly great. Over the last few years, they averaged FCF of S$30mn, which means that this pref share dividend alone will eat up S$12mn or 40%. There is one year the FCF went into -40mn, so if that happens again, the shares WILL plummet, both pref and common, like it or not. In short, not too much of a buffer that can guarantee payment until perpetuity.

4. The shares will trade on SGX come end Apr, which means that there might be downside risk if you need the money. It can be trading at $80, then you lose 20%. Of course I suspect it should trade above $100 in the next few weeks bcos Sg investors are suckers for such high yield. Already it's oversubscribed by 8x or so. But we must remember when the going gets vicious, it is likely this thing can go to $70 or even $60. The OCBC pref shares went to $80 during Lehman (now $104). If that happens, you have to ask if you are willing to add more :)

4. Hyflux is likely to buy back the shares asap, 6% of 200mn to pay every yr is no joke (S$12mn as mentioned, or roughly 40% of its FCF), but that's ok even if we just get a couple of yrs of 6% as the shareholders.

5. Finally if you subscribe, you need to do some guesswork, bcos you are not going to get everything. From the 8x oversubscribed no. probably $100k will get like $15k or something. Then again, you must be prepared you might get everything, but this is quite a low probability event now.

Conclusion: Probably an ok deal though not entirely secure like buying the safe and sound blue blue chips. Buyers would be betting on some implicit government support in really, really dire straits, a competent management that knows what they are doing and of course the 6% yield. 2 days left to go to the ATM to subscribe!