Sunday, July 23, 2017

A Prostitute's Ikigai



One of the points raised by the cruel commenters of Reddit when my article on why Singaporeans do not matter became a top read on this blog was that I have a a very narrow circle of friends who are gamers. I wanted to use this opportunity to show that my choice of social circle can lead to better insights compared to just hanging out with Singaporeans who want to take Instagrams of food or talk about their children's PSLE preparations.

Consider this scenario :

Someone wants to hire a Singaporean prostitute for a night of fun. When the time came to meet up with her, she did not show up, saying that she does not know whether she can get off her day job as a PA on time. The second time round, the person tried to make another appointment with her, but she was fussy and refused to service her client in a hotel room, instead she insisted that the client should do her outside (as in the open wilderness).

I felt that this is a teachable moment for my blog readers - maybe the prostitute really was in this line of work to attain self-actualisation.

Prostitution is a good threshold example when we look at different frameworks on career and life satisfaction.

The Western model of career planning we learn about in TED talks and self-help books is that we look for three things when deciding on a career :

  • Choose a career you are passionate about.
  • Choose a career that pays well.
  • Choose a career you are good at.

If we adopt the Western model, we can say that it is definitely possible for a prostitute to find that her job is a calling. In this case, the freelancer already has a day job and takes on clients because she enjoys the work, the work pays well and she is quite good at her job role ( Maybe she gives good GFE vibes and is always fully booked, who knows ?).

What's interesting is when you adopt the Eastern model and decide whether prostitution work will allow someone to find his / her "reason for being" or what the Japanese say Ikigai.

The Eastern model introduces a fourth element.

Prostitution must be the kind of work that the world needs.

At this stage, it is arguable as to whether Prostitution work can be someone's Ikigai.

A judgmental prude can argue that the world does not need Prostitution because it destroys families and introduces more negatives than positives to society. A more liberal interpretation can be that Prostitution can curb other sexual crimes by creating an outlet for men without conventional forms of sexual access.

Something that is acceptable by a society that emphasises individualism can be frowned upon when we assess it through the lens of a more communitarian society.

One thing is clear - Readers can benefit when they ask themselves whether their current vocation is their Ikigai.






Wednesday, July 19, 2017

Efficiently Inefficient #1 : How to think about market efficiency.

I am writing this article in Raffles City Food Court after a second day of lectures for the Part B Bar preparation exams. Two uncles sitting next to me are talking to each other about MLM and bragging about their recent trading successes via cryptocurrency trades. The conversation is quite fascinating but I have a blog article to write.

The book Efficiently Inefficient by Lasse Heje Pedersen was written as a textbook for folks who want an introduction into various hedge fund strategies. I thought the concepts may be be useful for retail investors who want to craft a more sophisticated investment strategy. After all, you guys are aware that I have since the last book, built a margin account which I intend to expand over the next two weeks. By August this year, about 4% of my total portfolio value would be leveraged by 200%.

We will start with a relatively simple concept of market efficiency.

The idea that markets are efficient is the idea that our stockmarket reflects all market information. Market prices always reflect fundamental value and readjust when breaking news occur. If markets are efficient, then active investing is pointless and your should just minimise your costs and just buy ETFs for your portfolio.

The converse of market efficiency is that the market prices do not reflect fundamental value and it is possible for active investors to succeed. Human beings make mistakes and it is possible for the investing crowd to become overexuberant or overly pessimistic.

The idea of markets being efficiently inefficient is a new one which tries to be a halfway house between market efficiency and market inefficiently. Markets are inefficient enough just to compensate money managers for their costs and fees but efficient enough to make it hard for a new market manager to enter the market.

If you adopt this belief, then market managers will typically perform well enough to be rewarded for the liquidity they provide to the markets but will find it challenging to perform beyond what they charge an investor in terms of trading costs and management expenses.

Some active managers will be able to exploit market anomalies over the short term but they will need to keep searching for new ideas to remain relevant to their clients.

I would this hypothesis to be correct. The only way to keep ahead of the markets is to keep reading and finding new ways to invest your money.

It does not make sense for anyone to be permanently financially independent without performing a porfolio review every now and then readjusting their investments to adapting to changing circumstances.


Saturday, July 15, 2017

Equity Management ( Last Episode ) : Time to move on and how to think about GLP.

I'm closing this segment in favour of something more reader-friendly next week based on more interesting hedge fund strategies.

I thought perhaps I'd like to show how the stuff I wrote on this column would be applied to the latest news on GLP.

Based on what can figure out on the news, GLP is being bought out at $3.38 and dividends declared in May will not reduce it's value which means that potentially whoever owns GLP will get to exit at $3.44 by latest April 2018. Looking at the current price, GLP is also trading at $3.29 which provides a nice 15 cent profit even for investors who decide to buy after the news has been declared.

Here are a few points :
  • Basic maths says that buying and holding GLP until next April will return around 4.55% or around 6.825% annualised. A lot of friends and bloggers have been sounding the alarm on such an arbitrage opportunity.
  • At this stage it is very easy to fall into a trap of equating this deal as a short term bond issue which gives 4.55% in 8 months. A 4.55% corporate bond is relatively stable but a buyout largely depends on whether the buyout will succeed. Serious losses can result in the buyout does not take place. 
  • At this stage, we can give benefit of the doubt to the folks who see this as a bond-like investment because the underlying mechanism is what we folks in law school know as a s210 Scheme of Arrangement which is welcomed by initiated by internal management. ( But a s210 can be shot down by the courts in a myriad of ways which will not be the subject of this article. )
  • So will you invest your money to earn an annualised 6.825% ? Depending on your level of risk aversion, some folks might. For me, I have much more attractive counters yielding 8-10% on my radar.
What is more interesting is when you apply leverage and try to mirror what some professional hedge fund managers do.

If you look at Maybank's lending rates for GLP, they consider this a Tier 1 stock so they only charge 2.88% for margin. Suppose you leverage at 300%, You can expect to earn over 6.825 * 3 - 2.88 * 2 or 14.715% annualised over 8 months.

14.715% is a decent return, but take note that if the buyout fails, you may be looking at perhaps 20% drop which would cost you 60% of your capital. There is no such thing as a free lunch.

This is something that I don't have the guts to do, but thinking about risk and return in this manner may provide more interesting insights that you can't find elsewhere.

( I am not vested in GLP )

Anyway, beginning next week, we will trying to go through the following book :











Tuesday, July 11, 2017

JD Aftermath #7 : Future of the legal industry

Image result for smu school of law

Today marks the end of my journey with SMU where I travelled back to campus for Commencement 2017. I wanted to save one last salvo before I begin preparations for the bar exams next week.

Today's highlight was Minister Shanmugam's speech. I would like to share some personal thoughts on what was presented to us today.

I felt that Minister Shanmugam's tone was generally positive. However, I can't help but feel that lawyers are going to be in for some really tough times unless they belong to the category that do bespoke work for big clients.

a) Elephant in the Room : No one knows how the legal landscape is going to be disrupted.

I interviewed with SMU as someone who has just left the data.gov team in IDA. At that time, I just completed my Coursera specialization on Data Analytics with R. I spent my time during the interview talking about legal analytics and ultimately how judgments can be predicted with the right deep learning algorithms. During my final year, some judgments from the European Court of Human Rights can already be predicted with 80% accuracy by just scanning the submissions.

In today's Commencement, what was not said was more important than what was said. Ho Kwon Ping's opening address briefly mentioned Fintech but nothing more was said about legal-tech.

IMHO, the Big Elephant in the room is that SMU does not have a module that teaches lawyers how to address the latest developments in computer science and give law students a sample of the skill-set required to master them. We are still not creating lawyers with the ability to speak to engineers and other technology professionals.

NUS, in contrast, offers this module to its students.

( If SMU wants to start, it should make Richard Susskind a compulsory read. Just sayin.... )

b) Inevitably, many of us will leave for greener pastures.

Minister Shanmugam puts up a series of very interesting slides which talk about how easy it is for other countries to take our lunch. In particular, Singapore is always under the threat of losing its position as an air hub. We also do not have a particularly efficient people-to-lawyer ratio compared to countries like South Korea (I felt that the Minister should not have withheld that ratios for the US but that's just me). If we want to succeed, we have to do more with less and a subtle message I received is that companies would not be satisfied with paying so much for non-bespoke legal advice. This is convincing to me because I spent over a decade in IT outsourcing.

I might be better off using the ITSM discipline to control a cadre of Indian lawyers to generate the low level research for more experienced partners before the passing the work to the client. I can even implement follow-the-sun where a document can be handed over to an Indian Call center and work complete before breakfast the following day.

If the legal landscape is going IT's way, then lawyers may be wiser to try to leverage on this outsourcing trend or at least find a bluer ocean to swim in.

c) Two great places to get into are Arbitration and Insolvency.

I felt a certain pang of regret when the Minister showed his info-graphic for Arbitration which was showing rampant growth.

Over a decade ago, fresh after my CFA I almost wanted to pick up a graduate diploma in Arbitration but was stunned when I was turned down for the inaugural batch of the course. Even in SMU, I pursued modules which taught me more about business so that I can be a better investor so I no longer had enough credit to study the arbitration module.

I was luckier because I spent a lot of time on Insolvency which was a tough subject. Initially I wanted to get into an area of the law where I get a leading indicator of whether the economy is coming to an inflection point. So I hope to get some real work on Insolvency in a later phase of my career.

Perhaps the most positive aspect of the talk is when the Minister sharing about growing areas of practice that rookies can get into.

What's next ?

Well what is next for me is that I will begin Part B next week.

I will withhold judgment until I meet the hordes of overseas graduates who are coming back to compete for a pie that growing smaller and smaller over time.

On the whole, I believe that SMU has prepared me well.

Heck, NUS Engineering and Finance school has prepared me well for this.

Bring it on !







Sunday, July 09, 2017

Why Singaporeans will not matter in the future.



I thought I'd spend Sunday writing about some of the key conclusions from the book Homo Deus by Yuval Harari within our local context. This is a very interesting read which I strongly recommend to all readers. In fact, you might be better off skipping Sapiens and going into Homo Deus straight because it talks about the future instead of the past.

a) Politically Singaporeans do not matter

If you look at the Oxley Road matter, one takeaway is that ordinary Singaporeans do not really matter in this unfolding drama, the support for our constituencies will not collapse overnight whether we preserve or demolish 38 Oxley Road. Ditto for the Elected Presidency.

The more interesting insight I gained from reading Homo Deus is the philosophical question of whether PAP is a political party or an algorithm. I find the algorithm argument convincing because the PAP is one of the most successful and consistent political parties in the world. Imagine the PAP as some sort of computer process or daemon with the Constitution and associated Statutes as some kind of database.

The PAP algorithm under the hood may work like this : Keep unemployment low within 5%, keep economic growth reasonable at 2-3%, manage constituencies to cover all basic and security needs. If all these conditions are being met, no opposition party can gain enough foothold to block a Constitutional amendment which further allows PAP to tweak the Constitution to continue to meeting these hard economic guidelines further entrenching their power.

Unlike the ministers that direct the PAP, the PAP is a highly intelligent algorithm without the requisite consciousness that knows Singaporeans better than Singaporeans themselves. Where Singaporeans can make a difference is only during rare cases like in 2011, when they gave an entire GRC to the Worker's party which triggers a flag for the PAP to make bigger changes to its social policies. But any intelligent programmer will know that code can be written to adjust to external sensor. This external sensor being an election poll.

b) Militarily Singaporeans do not matter

This is a much easier argument to extend because the nature of war itself is changing. A fat, overweight soldier can be a butcher in the battlefield because he can command a squad of attack robots. The consciousness of a solder is no longer an asset in a world where command and control can be simplified suing highly-refined AIs.

This will be a province of a few select Elite Singapore battle scholars.

c) Economically Singaporeans do not matter.

The ordinary Singaporean will have about 20-30 years head-start but will will become irrelevant much faster than his counterpart from another country because we are embracing a Smart Nation with a much bigger sense of urgency.

Look no further than the emergence of a robo-advisor, a low cost financial planner that would not push products down your throat or create a conflict of interest. The technology will not just render an army of financial professionals unemployed, it will create a better environment for millenials to invest their money. In the future, a robo-advisor might even be able to replicate active fund manager capabilities. I see this happening within 10 years.

As more technologies begin to disrupt the services, the population will split into a caste of "useless" citizens and a "super-elite" core that will control and direct an army of artificial intelligences to keep the economy humming.

I actually believe that the creation of the "useless' caste has already started.

I had a conversation with a bunch of friends the other day where 80% of the subject matter is about events in the virtual world  such as : What games to play. How to power level your Necromancer and what we bought in the latest Steam Sale.

A world in which Singaporeans do not matter may be welcomed by many Singaporeans

What I painted on this blog is very dystopian but I also believe that it would be welcomed by many Singaporeans who finally get the rest they desire. There will be no Marxist nightmare because the proletarian are too busy killing Diablo to overpower the bourgeois elite.

The ordinary "useless" class will withdraw from the marriage and breeding market ending years of heartache and frustration (Already happening based on the latest survey). They can plug into a virtual world where you can have a sense of achievement and even lead meaningful lives. (Unless you power-level and depend on someone else to give you free xp, which is akin to living on your inheritance). A world where non-degree males have problems finding employment, a government algorithm only needs to provide enough basic universal income to sustain a lifetime of plugging in playing games for an entire lifetime or watching Netflix videos.

The population will likely drop and the Singapore will be ruled by an enhanced and very small caste of High Programmers/Policy makers.

Your future Ikigai will be to reach 1000 paragon points.  

What can you do?

This is not a question that I will need to answer as I have already retired once. But this is a serious  challenge for my children.

The first answer for everyone is to simple embrace the future. For this to take place, some form of Universal basic income for internet bandwidth and food needs to be provided for the population. Efficiencies may make this a very small burden even for 5% of the population that pays taxes.

The next best answer is to attempt to join the super-elite. This is not going to be easy as the skills in the future would be advanced degrees in both biology and computer science. But perhaps 5-10% of the population will make the cut.

My answer is for folks within our generation to amass enough wealth to support just one or two generation ahead of ours. Investing in companies that will bring in this new age will maintain prosperity for our children who may or may not qualify to be super-elite.

If they fail, the family line will end with them but they will have comfortable lives. If they succeed, they will work very hard to keep the nation advancing into the future and may be able to continue going one perhaps a few generation more.






Saturday, July 08, 2017

Equity Management #19 : Dividends strategy versus Deep Value Investing.

During my preparations for the talks at the end of the month ( which is rapidly selling out ! ), I would go to Lee Kong Chian Library and take down some interesting observations on the market from the Bloomberg terminals.

Today's observation goes beyond the scope of my presentation in July so it's better to share this on my blog for intermediate investors. It will also give attendees and idea of what my presentation would be like at the end of the month.

An important question investors are curious to know is whether deep value strategies outperform a well-crafted dividends strategy. I have always maintained in my talks that a dividend strategy is, in many cases, inferior to the deep value strategy.

( Information on my dividends strategy will be shared on my talk itself but let's look at simple deep value strategy I back-tested on Bloomberg.)

Suppose we take the bottom 20% P/E ratios from a set of local SGX stocks. And then within this set, we find the bottom 20% P/B ratio. We will end up with a list of deep value stocks that belongs to the province of deep value investors.

Backtested 10 years and annually rebalanced, the performance is abysmal :Returns are 3.07% and the portfolio is highly volatile with a semivariance of 23.40%.

Investing in the STI ETF would have gotten better results.

As I could not believe my eyes, I then tried to observe the stocks selected by this approach. The market returned mostly Chinese companies.

So the next step would be to limit the scope of this strategy to locally domiciled companies.

The improvement is very dramatic - Once you filter out the Chinese companies, the returns become 52.46% annually ! The semivariance becomes higher at 34.69%.

This is superior to any dividends strategy I will be presenting end of this month and possibly superior to any performance of active managers covering the Singapore market !

Some preliminary suggestions for investors :

a) If you go for deep value, filter out the China stocks in SGX.
b) While your returns will improve, your risks will also be magnified.
c) The backtest does not account for the lack of liquidity of good deep value stock counters. So you might even be able to buy the stocks suggested by the screen becuase no one is selling them.
d) While it's tempting to leverage such counters, brokers are unlikely to give you decent lending rates for deep value stocks.
e) Unlike dividend stocks, deep value counters may not occasionally give you a reward you for holding counters over time.
f) No release of dopamine in your brain when you get rewarded with dividends. ( Which is what got me hooked more than a decade ago ! Dividends are the opiate for the investing masses ! )


Wednesday, July 05, 2017

Personal Update : The Road to Commencement and Part B.

Just another personal update as I have just returned from my short holiday which has been spent mostly queuing for Universal Studio rides. 

The next few weeks is going to be rough :

a) Signing up for a 2-day shorthand course

Fans of my blog know that when it comes to lifelong learning, I tend to go all-out and take a skills mismatch quite personally. One of the problems I faced during my internship is the taking of Court Attendance notes during Pre-Trial Conferences. My fingers were too slow when taking the notes down and the quality of my work was generally quite poor. ( These are verbatim notes, not the summaries which I can comfortably handle in the past. ) 

So yesterday, I shortlisted a shorthand approach called the Pitman approach which is efficient but has a high learning curve and will begin lessons this Friday. I am rushing this through so that I can unleash this new Talent Tree/Feat when I begin my lectures for Part B. 

I will let everyone know whether I can claim my Skillsfuture credit for this class.

b) Cryptocurrency classes on Coursera
 
Just one day before Commencement, a Coursera class on Blockchains would be opening up next Monday. I am now considering whether I should attend this course. The question is whether I should pay for the course to do the capstone project. 

This is somewhat risky since Part B has a distinction grade and a senior lawyer half-jokingly told me that I should not try to merely try to just pass my bar exams. ( Gimme a break I SMU is brutal enough ! )

If you do wanna sign up for the course and want a study buddy, do let me know. 

c) My next talk  "50 Shades of Dividends Investing"  

This talk is shaping up well and all my content has been incorporated into my slides so I can share my overview here. Generally speaking, expect a highly quantitatively driven talk. I don't make bald assertions without backing it up with a back-test from a Bloomberg terminal. You can sign up for this mega-event here.

Some of my findings should be counter-intuitive and controversial enough even for intermediate investors but there is definitely two useful investment strategies which I will share in my talk.

The Overview look like this :

The Corporate workplace as a Brutal Master
Fifty Shades Darker
Earning
Saving
Investing your way to freedom
Investing specifics :
The flaws of the STI ETF strategy.
How dividends can assist you.
Importance of REITs in a dividends strategy.
Two cool approaches to dividends investing.
Question of leverage
Fifty Shades Freed
Attaining Financial Independence