Tuesday, January 29, 2019

The Model Thinker #5 : The Normal Distribution

Image result for normal distribution

The normal distribution curve can be used to assign probabilities to values. Every data point can be condensed into two numbers, the mean of the curve which measures the average, and the variance, which measures the dispersion of the curve.

The normal distribution curve is symmetric about their mean and finds application in many areas of nature. The average height of the test scores, size of flora and fauna, life spans can all be modelled by the normal distribution curve. A very important theorem is the central limit theorem, where adding or averaging the random variables produces a normal distribution curve.

Are there different kinds of distributions ?

When random variables are added together, the resulting distribution curve is likely to be normal. But when random variables are multiplied together, a different distribution curve results. This is a lognormal distribution.

Image result for lognormal distribution

Salary raises are typically expressed as a percentage increment. So folks with higher salaries receive a higher increment for the same percentage raise.Because of this, the income of population tends to follow a lognormal distribution curve. Notice that larger variances in a lognormal distribution curve tends to result in fatter tails to the right.

The lognormal distribution possibly provides a better explanation as to why inequality in our society exists - increments are expressed in percentages so higher paid professionals are given higher increments. Things would be very different is increments are expressed in absolute figures - we might even end up with a gentler form of capitalism.

Financial professionals often have no choice but to express market returns as being normally distributed. When they have to work with actual stock market levels, they apply a logarithmic function before feeding it into a regression exercise.

The truth is that most of us are aware that markets generally do not follow a normal distribution, but we are often forced to because more complicated distribution functions ( possibly suggested by Nichola Taleb ) may not even have a means of defining the variance of the distribution.

You can read about how the abuse of a mathematical function was probably what almost killed the markets in 2007 here.






Sunday, January 27, 2019

On employee engagement, productivity, and the death of a Singapore Son.

One of my favourite bloggers, TACOMOB has not only made a come back after taking a long break from his blog, he showed up for my preview this week. I strongly urge you guys to read his comeback article here because it provides a lot of food for thought.

The FIRE movement is strengthened by the dissatisfaction of workers in Singapore. While we are not responsible for the poor work cultures in Singapore that lead to so many workers becoming disengaged, we derive a lot of our popularity from it.

Attaining FIRE often requires massive personal sacrifices, sometimes we spend less than 20% of our take home pay. The only impetus that can lead to such behaviour is immense dissatisfaction at work and the dysfunctional workplaces that we have.

Attached is a word cloud spontaneously generated by the preview attendees when I asked them to use one word to describe their working life.


On TACOMOB's blog article, he did not explore why we are disengaged. I thought I'd explain it in a politically incorrrect manner that, if he had done so this way, he would have gotten into trouble in thin-skinned Singapore. So it is best that a Singaporean blogger do this.

The brutal truth is this : Singapore managers, unlike their Ang Moh counterparts, are culturally immature.

Another words, Singapore managers suck.

The Singapore corporate landscape is brutal for a developed economy. Singapore workers work 45 hours a week. German workers work on about 26 hours a week. In spite of that, Germany remains one of the most successful economies in the Eurozone today.

I personally transitioned from an American corporate culture to a Hindi-speaking start-up and then out of the private sector into public organisations run by locals. I rate my experience with expat managers like TACOMOB as the most pleasant and egalitarian I ever experienced in my life. I rate my experience with local managers as one of the most unpleasant, actually worse than my time in an Indian start-up. I still remember the days when a director of mine, a control freak, would expect me to parrot his exact phrases when preparing me for a presentation to a room of senior directors. ( Even though I am way better a public speaker than him. )

Singaporean bosses are "ngiao" and very stingy with praise. I wrote a letter thanking a committee for letting me play the role of secretariat and was subtly told that I should not have done so (without reason). They are actually very entitled and are protected by our resistance to a welfare state economy. On the other hand European and US managers are trained to deal with very aggressive unions. How many local bosses ever had to negotiate with an angry union leader or even an employee who does not need his job to feed his family ?

What is the root cause of this ?

The time it takes to build a corporate culture is much longer than the time it takes for a country to leap from third world to first world. We've only had half a century of independence, so Singapore has the hardware of a first world country but we are essentially a country running third world management software. If you observe the behaviour of a top senior director in a government agency, his management methods assume that his men are going to be lazy and will ring-fence their work - sad thing is that he right, these agencies often retain the workers who are the best at doing that.

The sum total of our immature corporate culture has taken its toll on the country lately.

We've lost a prominent son of Singapore to SAF training lately. Whichever side you are in, safety in a dangerous organization like the SAF can never succeed if it is top down. It has to be maintained on the ground by the rank and file. But for safety to be bottom up, a soldier must feel that he can report a lapse without facing consequences.

In my penultimate ICT with the SAF I was in a Brigade and during an exercise, my land rover's engine stopped working in the middle of an expressway. After we pushed the rover to the side, The MT line sent a tow vehicle to our location but left us at the side of the road. In the end, we called our Signals CSM who is a cab driver to pick us up in his cab. When we finally got back, the canteen was closed and we basically ended up skipping one meal.

I wanted to complain, but there are so many impediments. First of all, no one died. Subsequently, a feedback session recorded my feedback, and there was a promise was made to update us on the next ICT. That did not occur. By next ICT, I did not want to pursue the matter because it would brand me a "troublemaker" (I was not legally trained then). As much I as liked my commander, brigade command staff were almost all chosen from powerful echelons of Singapore society. Offending them or giving them extra paperwork may have long term consequences.

After what happened to Aloysius Pang, I felt a deep sense of regret of my conduct during my days as an NSman.

The right thing to do would be to hound the powers that be until we get some assurance that land-rovers are properly maintained. I might be punished - the biggest punishment would be relegated to a  Battalion, but maybe if some other soldier were more disagreeable, maybe, just maybe Aloysius Pang would be ok today.

Unlike other folks, I'm not calling for the blood of the CDF or the COA. Once we become NSmen, we have more autonomy and power in society. Maybe some professionals who can write nasty letters and have plenty of investments can summon the courage to take a stand against safety lapse.

It is up to us, professionals, possibly FIRErs who might be financially independent, who can pick a fight and not get squashed by the military machine.















Friday, January 25, 2019

Should you buy REITs based on a "Good" Sponsor ? Part 2

Here's a nasty thought experiment.

Suppose on one hand, you have a team that is a sample of RI, ACS and Hwa Chong students. Their average PSLE score is 265.  We shall call this team A.

Team B consists of a bunch of kids from government secondary schools, such Changkat Changi Secondary or St Gabriel secondary secondary  etc... Suppose this selected sample would also have an average PSLE score of 265 ( this is a small sample size and some kids volunteer to go to ordinary schools ! )

Is there a significance in the life outcomes of these different group of students ?

This experiment was done on the outcomes of SAT candidates who enter different universities in the US and it was found that, generally speaking, the SAT score is a better predictor of wealth and success in life than the quality and brand name of the school itself.

Now suppose Singaporeans get to bid to buy a share of the earnings of this sample of students in the future. I think, given how kiasu we are, the RI boys will command a much higher price.

But because you will be paying less to get the cash flow from Team B, you should expect the folks who bet on Team B to make more money moving forward.

The last time we had a discussion about the debunking the "Good Sponsor" theory, we had a pretty good debate because I understand why some fairly experienced REITs investors got upset over my article.

Today I'm going to share with you a different experiment that basically came  up with the same conclusion.

Another well known anomaly in US is that stocks with a lower average volume tend to outperform. This can happen for many reasons. Perhaps low volume stocks present an illiquidity premium. It is also possible that low volume counters receive lower analyst attention so it may be amenable to some bargain hunting.

For the baseline, equal weighting REITs for 10 years ending on 31 December 2018 would result in returns of 18.45% with a semivariance 10.36%. Do note that I've been promoting an equally-weighted all-REIT portfolio for quite a while in my previews.

So once again, I decided to conduct an experiment, this time using the factor of average volume of the REIT over the past 6 months.

Here is a snap-shot of the ten REITs with the highest 6-month average volume as of two days ago.


So once again, we are looking at the more popular REITs in the stock market. With the exception of Lippo, any sponsor of these REITs can be subjectively agreed to be "good sponsors". I say this because I'm still not getting a firmly agreed stance as to what is a "good sponsor".

What about the ten REITs with the lowest average volume ? Well it looks like this :


The reputation of this list of REITs is not definitely as stellar as the previous screen-dump.

If you have been buying the REITs with the highest 6-month average volume for 10 years right up till 31 December 2018, you would have done well : 19.24% returns with a semi-variance of 11.55%. The problem is that you are getting higher returns for a higher downside risk.

Here's the magical finding.

If you have bought instead the REITs with the lowest 6-month average volume, you would return 20.28% with a semi-variance of 9.95%. The REITs with the lowest volumes produced a better return at a lower risk.

Here's my proposed conclusion and you are free to argue a different interpretation of these results.

Choosing a REIT with a lower average volume is a superior investment strategy to choosing a REIT with a higher average volume. To exploit this advantage, you have to do the opposite of what many investment experts and value investors say : you have to avoid or to reduce your exposure to REITs with "good sponsors". This is because these good sponsor manage their REITs so well that retirees and non-experts plough their money into these counters in the hopes of getting safe yields, many of these guys just want 6% yields.

The collective retail investor fund flows into these better run REITs counters makes it harder to derive extraordinary profit from them.
















Tuesday, January 22, 2019

The Model Thinker #4 : Modelling Human Actors

Image result for vitrivius man

The first serious chapter on modelling covers human actors.

There are two general approaches to modelling a human actor.

a) The Rational Actor Model

Modelling human actors as being rational has come under a lot of criticism but no other model has the kind of consistency and elegance offered by this approach. By making an assumption that a person will act rationally, you can also come up with behaviour that approximates that of a highly intelligent player.

In the Rational Actor model, individuals choose the action that maximises their payoff. This pre-supposes that a payoff function is defined over a set of possible actions. If an investor is rational, between an investment that generates 10% and 12% but with the same semi-variance, the investor will choose the latter strategy.

Upon further inspection, you will realise that most human beings are hardly rational. One component of rationality is that ability to consider all options. All alternatives must have a preference ordering. Another aspect of rationality is transitiveness. If you like A over B, and B over C. You must like A over C. ( I am hardly rational when it comes to movies )

If you come into the financial markets assuming that all market actors are rational, you will be in serious trouble. Even I have serious quirk which I am not really willing to show the world even though Dr Wealth colleagues have asked me to give more details on my portfolio - one of the quirks I am less proud off is that I like to invest in denominations of eight... For example, I have 8800 shares of UMS.

To blunt the lofty assumptions of Rational Actor theory, economists now add theories of loss aversion and hyperbolic discounting into whatever models they have. We take larger risks when faced with losses, and we apply a stronger discount rate of our immediate future.

b) Rule-based model

Rule based models do not use a payoff function but assumes that people will behave based on specific behaviours. One possible way to model the behaviour of an average Singaporean is that, based on government  data, he is likely to set aside 20% of his take home pay.

The rules-based model may be fixed, but common sense tells us that people adapt to policy changes and changes. When the salary of a person goes up, he might be able to save a larger proportion of his income. He may have to forego savings when he gets retrenched.

The spirit of modelling is to adopt the approach that all models are wrong one way or another. One criticism of modelling the behaviour of human beings is that changes in policy or the environment will produces behavioural responses by those affected. This renders past behaviour inaccurate.

But we know this problem well in quantitative investing.

This is the problem of non-stationarity.


Sunday, January 20, 2019

When financial bloggers take a break

Image result for i quit wwe

This is a pretty rough week.

First 15WW has expressed a desire to take a break. You can read about it here. Then yesterday I received news that Cheerful Egg would also be taking a 90 day break to reassess his priorities.

These are good bloggers to follow because their material is quite candid and original.

The work of a financial blogger is a thankless job. A few years ago, a lot of us came in with the idea of monetisation of our blog material. That turned out to be much harder than we thought. Even if we can cultivate a legion of fans, unless we conduct courses, give talks or make physical appearances, the profits from writing articles can only float to maybe a few top blogs in Singapore. If you think about a successful local blogs, maybe only Investment Moats and AK71 come into mind.

Once you cannot monetise a blog, the disadvantages kick in. Unless your blog is completely informational, any amount of candour can potentially affect your day job and employment chances. I constantly find that everything I write can offend potential employers. Luckily, I can do it because I don't really need an employer these days. Imagine how hard is it for someone trying to FIRE at the same time.

Here are some for the consideration of other financial bloggers :

a) Read a lot. 

The crucial point about sustainability is to read a lot and reflect your reads on your blog. Your articles are novel because you adapt what you read into living in Singapore. You would also have a constant stream of articles that will eventually broaden your fan base.

For me, thought leadership matters a lot so when a new book comes online, I pay more to get it into my hands so that I can incorporate it into my training materials or blog.

If you reason that reading should be factor into the time I spend being a trainer, then I actually work harder than most office workers in my current life. I don't include this tine in my life-energy calculations because I enjoy reading and would do it even if I don't have classes to teach.

b) Don't parrot mainstream ideas unless you have a novel interpretation of materials at hand.

I thank SMU for this idea. When grading presentations, an A grade is only reserved for a "novel" interpretation of legal ideas.

The relative popularity of Marie Kondo has led me to clear about 40% of my clothes and surrender it to H&M for a few vouchers. The problem is that in becoming mainstream, this form of minimalism is less interesting to write about.

What is more interesting is the question of which investments "spark joy". If I can come out with a thesis of why Asia Pay TV sparks joy if it is below 13cts, I have a nice article that can pick up a larger share of the attention. 

c) A blog is only one piece of your personal brand management.

As more sophisticated blogs come online and some of us start to align with very powerful groups like Seedly, the question is what kind of room does a blog simpliciter have in the grander scheme of things.

Seedly has become a behemoth in the industry. I think a blogger ignores them to their own detriment. This year, I realised that it's way better to cooperate and see if there are win-win initiatives with them.

This is going to be hard for introverts but bloggers need to put themselves out there to change the game. This means giving talks, making physical appearances and conducting courses.

I have been thinking about the question of using podcasts and vlogs to extend my reach myself.

d) A financial blogger needs to be a little selfish too.

Of course bloggers have to address what they can do for an audience, but a more important question is what can the audience do for you. Many readers just scour blogs for stock tips.

The surest path to burn-out is to keep giving but not getting anything out of your blogging activities. Your blog should be about self-improvement and even some degree of self-aggrandisement.

I can keep doing this because I don't really care about the audience -  I know folks want a brutal take on polytechnic students, why they need to be careful when signing up for a private degree, and definitely how our CPF program is like schizophrenic prostitute. These postings give me 1000 views within a day. It took me only 25 minutes to write the CPF Prostitute article.

But I also write for myself to chronicle key ideas to make me a better investor, Articles on Aristotelian categories would be lucky to land 200 hits in one day. These articles are harder to write because all these ideas need to be shoe-horned to practical life. In the future, these are my hidden real options to publish a full thesis on investing.

I think more bloggers will quit  or at least take a break in 2019.

But even if a lot more bloggers will take their place, I always feel a little sad when this happens.





Thursday, January 17, 2019

An Earnest Conversation - Dealing with posts by commissioned sales professionals that you may disagree with

Image result for eviscerator

Money Maverick wrote a really sweet post which showed his human side.

I would like to invite everyone to read it here.

While you don't have to agree with Money Maverick (I certainly don't), you will probably accept that he should be treated like a fellow human being and did not deserve this level of abuse online.

Not everyone in the community likes the style used by Money Maverick, some folks say that he's too slick and basically would share details on the securities he pitches only to his clients. As a trainer who does uses my blog to promote my personal brand, I can understand that Money Maverick has to keep doing what he does because one successful case conversion can net him a reasonable amount of commissions and allow him to attain his life's aspirations.

So today, I just want to showcase a better way to deal with such articles.

A lot of unhappiness was directed at his claim that he has a fund that can provide passive income of over 7.8% but at a volatility below REITs.

a) Instead of attacking him for being secretive, why not try to make an intelligent guess as to what it is ?

If I may guess what he's referring to, it may be some kind of fund that specialises in junk bonds. As 7.8% yield is rather high, it may be focused on specific region like Asia. The low volatility probably comes from the fact that these bonds are thinly traded anyway and you might be taking on a larger share of credit risk. There is also forex risk as high yielding bonds come from countries experiencing high inflation. So while traditional measures of volatility would be rather low, you might not like it when one of these junk bonds default.

But why not challenge yourself to take a higher road ?

b) Another approach is to see if you can beat his fund at his own game using DIY components. 

We can find REITs that yield over 7.5% from the REITDATA.COM, simply picking out these REITs will give us about 10+ counters. I did an average of these yields and I got 8.61% which is already higher than 7.8%.

[ In November 2018, I backtested a dividends strategy with REITs and it returned about 19.8% a year with a semivariance of only 11.17%.  Volatility of REITs is already pretty low. ]

The next question is how do we get volatility that is below that of REITs ?

Simple. Just add retail bonds.

If we can add Astrea IV Bonds into the mix at a proportion of about 18%, you will end up with a portfolio that gives out about 7.8% dividend yields but have, generally speaking, lower volatility than REITs.

( A craftier investor can add a commodity ETF because of the lower correlations )

So if I have $100,000, the appropriate response is to structure a portfolio of equal weighted REITs that yield at least 7.5%. In such a case, $82,000 should be invested in this portfolio. $18,000 should be used to buy Astrea IV bonds.

So with a simple exercise, I can confidently build a portfolio that has a decent probability of beating Money Maverick's proposed fund.

If you are willing to go DIY, you have a huge edge over solutions offered by sales professionals - You pay lower commissions, management expenses and there might even be tax advantages as we're talking about S-REITS here.

Let's try to make the financial blogosphere a nicer place for folks like Money Maverick. There is no need to condemn him for his choice of profession.

When he proposes an investment idea, challenge yourself to guess what it is.

Then take the high road and engineer a DIY solution to the best of your ability.

This why I always take the position that blogs like Money Maverick can potentially teach us a lot about investing.










Tuesday, January 15, 2019

The Model Thinker #3 : Aristotle's categories

I can argue that Aristotle is the philosopher who inspired all of quantitative investing. Even before the establishment of modern financial markets.

Aristotle created a framework of inquiry that is relevant even today, his ten categories being : Substance, Quantity, Qualification, Relative, Where, When, Being-in-position, Having, Doing and Affection. By inventing these categories, it is possible to classify everything under the sun. Once a set of objects can be categorised, we can start hypothesising about the relationship between these objects.

Naturally, in this modern world driven by data, a good working theory can result in extraordinary profits.

To bring this discussion down to earth, I have been experimenting with folks who show up for my previews. As my quantitative models have been working pretty as of late, I want to see whether the same quantitative discipline can be used to learn more about folks who are passionate about learning about investment. This is an example of one-to-many thinking : can I use the same models I learnt in my investing work in my new job as a trainer ?

The following slide represents my maiden attempts to understand my preview participants better :


I tried to distill the MBTI personality tests into one Mentimeter slide and attempted to figure out what "personality" do preview participants have.

After some crowdsourcing, I have at least one data-point that folks who show up for previews have the ISTJ personality. Introverted, data-driven, and decisive thinkers who otherwise would make great administrators and accountants.

So I can now imagine my course preview as some kind of equity screen that returns participants who belong to the ISTJ personality category.

Of course, just because you have an equity screen with a selected stock does not mean that you can you can profit from it.

Being an ESTJ-turn-ENTJ, communicating with ISTJs needs to be picked up as a skill.

My next step is to subtly change my materials to reflect the needs of my audience. Every assertion should be backed by some data. Every major outcome should come with a series of steps on how to achieve that.













Sunday, January 13, 2019

What does the Finance Industry and the Social Development Unit have in common ?

Image result for stephen chow ugly girl

Until the Millenials developed Tinder, dating was not a straight-forward process for guys in my generation. An entire generation of men were emotionally scarred by the Social Development Unit or SDU.

I was, fortunately, not a victim of this government initiative. Instead, after being burnt by a match-making attempt by my parents, I developed my own quantitative approach towards meeting women.

( This is why I have a family today in spite of being D&D otaku and troll. That story is something I will tell another day. )

Some of my friends were not so lucky as to escape the clutches of SDU.

Seduced by cheap government-sponsored dating activities, many of my friends fell into the trap of attending SDU events, some retained their sanity long enough to tell me their tales of woe. Through the horrors experienced by my friends, I have heard tales of people so unattractive, boring or so disagreeable, they could have been tales to scare naughty children being putting them to bed.

It took me and my friends ages to realise what the ultimate objective of SDU was.

In my opinion, it was clearly not an attempt to match-make singles.

The aim of SDU is to hurt your self-esteem and lower your standards so drastically that you would eventually settle for anything that comes by in your life. Imagine if you are fed army rations everyday. One day you are given a slice of kaya bread - that would be the best meal you ever had in recent memory.

Thanks to Tinder and Coffee Meets Bagels, the SDU is now a relic of the past.

But very recently, I have noticed that the finance industry is now using the same tactics to increase their profits.

Just the other day I went a new eatery in Selegie Road and I was lucky enough to overhear a conversation between the proprietor and his old NS pals. The proprietor worked in insurance and was able to raise enough capital to start the food establishment with another insurance buddy of his.

He shared a few interesting snippets about the industry. Namely (a) closing one case nets him about $1,600. (b) He can make about $6,000 - $7,000 a month (c) Even without closing cases, his previous wins can give him about $3,000 - $5,000 a month. Payments seem to stretch years (d) He does this by pure cold calling and he does not sell to friends or relatives.

I came away with the impression that making it in insurance is all about dealing with rejection, but more importantly, the commissions earned has to come from someone - the hapless insurance consumer. If closing a case nets the agent $1,600, it basically means that his client loses about $1,600 when he agrees to whatever advice was given to him.

How does the industry collaborate to lower the standards of consumers since so much money goes into paying commissions ?

Broadcasting lower returns, of course. Of recent note, I have seen a lot of insurance ads talking about the returns a consumer can expect from their products with an investing component. These numbers tend to hover around 2-3%.

Broadcasting these returns has the same effect as SDU, eventually creating a lower anchor for hapless investors to lock onto. Whoever manages these investment products have greater leeway to provide the higher commissions to insurance agents, who can, in turn, make more aggressive moves against consumers.

Here's how can reverse this anchor.

After a brutal 2018, the SPDRs STI ETF still returned about 6.5% since inception, or more than 8% if you consider its performance for the past 10 years. If you take the Nikko AM Investment Grade Bond ETF, you can expect a yield of 3.2%. A combination of these two ETFs will still net yourself more than 5%.

Another argument is that we can expect inflation to be around 1.9% every year. Even though Singapore has experienced lower inflation than other OECD countries, you are in essence locking yourself down in an instrument for real returns at around 1+% at most. One way to troll your advisor is to ask him to express his projections in real instead of nominal terms.

If the insurance agent insists that the lower returns are guaranteed, make sure you read the small print. There may be a loss of liquidity for a number of years. In any case, directly pumping into your CPF-SA will net you a risk-less 4%. The fact that some products lock you in for a number of years and yet the returns are benchmarked against SSBs which do not have a lock in period is possibly the result of very sneaky marketing.

You should also remind your agent that capital guarantees are predicated on the insurer not falling apart. (AIA was bailed out by the US government during the 2007 recession )

The financial industry is creating the same trap that has hurt an entire generation of men who have attended SDU dating events in the past. Don't let them trap you into accepting illiquid investment products for real returns of less than 2%. Instead of going SDU, why not venture overseas like Vietnam and Thailand to meet the women there ?

Similarly, separate your investment and insurance needs.

Where possible use a DIY portal for your insurance needs. Pay a small premium for pure insurance products, minimising commission payments, then invest the rest in a portfolio of standard vanilla equity and bond ETFs.



Friday, January 11, 2019

How to take a crap on quantitative backtesting.

With the efforts I have been gaining to show up some misconceptions about investing in local stocks, I would have thought that I should be become by now public enemy number 1 from value investors. Sadly, I have read some attempts to discredit quantitative backtesting and I find that they are either holding back key information, or generally lack intellectual sophistication.

As such, I will take on the burden of criticising quantitative backtesting myself. This is because, of all people, I cannot afford to drink my own Kool-aid. Not only could I stand to lose money if my methodology is flawed, it can affect the portfolios of my students too.

There are, in fact, three valid responses when faced with quantitative backtesting data.

The first, which I support, is to simply accept that is a superior approach towards investing. The second is to disagree and argue that you can get superior returns because the model inadequately capture all the risks associated with this strategy. ( For example, the strategy of small stocks is plagued with the problem of illiquidity. ) The third is to simply argue that backtest models model the past and the strategy simply will not work in the future. You need to have a better way to make money if you argue as such.

A good quantitative strategist must address three specific issues :

a) Data Mining

If you spend enough time interrogating data, you will always find a combination of factors that can result in excellent past performance. There are hundreds of factors to play with, all the quant needs to do is to set a time range and start grinding the numbers through the factors and pick out the best one of the lot. The problem is always the question of whether the strategy works in the future.

The only way I address the problem of data mining is to use only tested factors in the US and repeat the experiment in SGX. Then I do some work to ensure that I'm not crazy by testing out the strategy in Bursa Malaysia. I can safely say that dividends sustainable by free cash flow has not only passed all the tests but also receive support from value investors.

b) Non-stationarity

Imagine you have a bag of red and white balls. You draw out the balls one by one. Over time you have a certain idea of the proportion of red balls to white balls after a while. Sadly the markets do not work like this - when you invest in the markets, the markets will change the bag every now and then. So when you are having the Great Recession and lowering interest rates, you are drawing from a different bad than when you are facing the trade war between China and the US.

This is a much harder issue to address as this means that the returns and standard deviation can change as we enter a new stage of history.

This is not a easy issue to address. My only defence is that if there is a significant change in history, it should be argued as such. We had the same PAP government for 50 years. If anything, Singapore investors have a greater reason to stick to models captured in the past than, say, Turkish investors.

My only defence is to keep expanding the backtest timelines and watch other markets when strategies begin to fail. Of late, the small firm effect is not as profitable as before and value has been losing to growth for over a decade in the US.

c) Model misspecification

This is so hard, I struggle to understand how to even deal with it. When we fail to select factors when we create a model, a factor like the low-variance effect can either be interpreted as a market beating anomaly or a risk premium. So the model is incomplete.Academic researchers have found 316 different factors in quantitative investing.

 The combination of factors I employ are the obvious ones documented by texts. The sample size of Singapore markets is also very small such that using more than 2 factors at one go will result in a set of stocks that may be too small for diversification.

So I have chosen to ignore this critique of quantitative investing.

Maybe a professional quant can share some ideas on how to address these issue when you run a professional fund.









Wednesday, January 09, 2019

The Model Thinker #2 : Seven uses of models

Image result for seven supermodels

In this second installation of The Model Thinker, I will talk about seven uses of models.

The author uses the REDCAPE acronym to illustrate these seven uses which I will illustrate here for folks interested in their personal finances :

a) Reason - To identify conditions and deduce logical implications

One of the best illustrations of this usage is Ray Dalio's Big Debt Crises he showcases one way to look at how accumulating foreign debt can lead to a depression later on. A detailed discussion of this model, unfortunately, cannot be done on a blog. We do know that taking on a lot of foreign debt can led to tragedy later on.

b) Explain - To provide testable explanations for empirical phenomenon

I attempted to explain the underperformance of a REIT portfolio due to an investor's preference for REITs run by 'good' sponsors using quantitative backtesting. This racked up a lot more unhappiness than I originally intended, but I will possibly do up a sequel to that article quite soon.

c) Design - To choose features of institutions, policies and rules.

The CPF Life program defaults to the Standard Plan which is drawn from the concept of choice architecture illustrated in the book Nudge by Richard Thaler. The government obviously wants more people to load up on annuities in their retirement plan so a sneaky way to do it is to make it the default. This is excellent policy design, but I fully intend to switch to the Basic Plan the moment I get to do that.

d) Communicate - To relate knowledge and understanding.

If you follow my block, you might know that I am getting rather obsessed with building a taxonomy/epistemology of financial knowledge. I was inspired by the TCP/IP stack and a writeup from a finance professor. I hope to explain the problem of financial knowledge better. If I fail, at least the IT guys and engineers will appreciate my use of this model.

e) Act - To guide policy choices and strategic actions.

A very simplistic way to locate Singapore's position in the market cycle is to look at trends of CDP growth, inflation and unemployment. Once we can guess at our market cycle positioning, we can start to determine our asset allocation. At this point of time, we might be in a contracting phase - this means buying bonds in spite of their lack lustre returns over recent years.

f) Predict - To make numerical and categorical predictions of future and unknown phenomenon

What does it mean when my strategy of buying stocks yielding more than 4% and having a P/E less than 12 returns over 20% ? A lot of folks misinterpret this as  projection of future returns of 20+%. Actually, I build quantitative models to maximise the chance of outperforming the market average. So I am predicting that this 4% yielding and PE would do better than holding a basket of Singapore stocks in equal weights - a less dramatic prediction than +20% gains every year.

g) Explore - To investigate possibilities and hypotheticals.

Over the next few days, I will be investigating whether REIT's with 'good' sponsors have a larger average volume over the past 6 months.  I hope to use these insights to see whether I can put a second nail into the "Good REIT Sponsor" idea that many investors currently have.

In summary, models are very powerful tools to assist an investor. Hopefully as I explore this book further, I will learn of new ways to structure an investment plan.

Sunday, January 06, 2019

Your CPF is like a Schizophrenic Prostitute.

Image result for female two-face


The inspiration for this article started when I suggested to another financial blogger to write a CPF article and pepper it with sexual references to push readership figures. He then said that his brain is too tired, so I challenged myself to write the most pornographic CPF article ever written in the financial blogosphere. If you get triggered too easily, you can always tune out of this blog and go watch a Hallmark Christmas movie instead.

a) Investing in CPF can be like visiting a special kind of prostitute.

If investors were Johns, I expect buying stocks is like visiting prostitutes who are good looking and generally give good Girl-friend Effect (GFE). The problem with equities as prostitutes is that in a particularly bad year like 2018, instead of giving the John a good time, the prostitute climbs on top of him and shits all over face. I have even learnt the hard way that stocks like APTV will even bite off your nipples off in the process.

Enter the CPF.

The CPF is the kind of prostitute that gives a consistent guaranteed performance. 2.5% guaranteed is not bad, but it is the kind of transactional relationship with little GFE. She barely gets the job done. Even the 1% benefit only applies to the first $60,000 of your assets adding only an extra $600 every year.

Thus, the popularity of the CPF program is this consistent performance.

Sadly fixating on 2.5% allows financial institutions to sell riskier programs with returns of 3+% and then tout this investment as a better alternative to fixed deposit. A fixed deposit is not even visiting a prostitute, it is risk-less and guaranteed by SDIC - this is more like an innocent date with platonic lady friend.

Worse, financial institutions selling low return offerings also condition investors to expect bad service in Geylang.

b) The CPF-OA account is schizophrenic prostitute 

The second point to note is that your CPF-OA account has multiple purposes. This is the biggest flaw and biggest strength of the program. Imagine hiring a prostitute not just to give you a good time, she will also do your housework as well.

This is the biggest watch-out of the CPF program.

If you use your CPF to buy a larger home, expect to less retirement adequacy. A prostitute cannot give you a good time and then go wash all your dirty dishes at your sink at the same time. You will not get good service.

Buy a 4 room flat and you may be able to replace 75% of your income when CPF Life kicks in for you. Buy a 5 room flat and your income replacement may dip below 50%.

c) To resolve this prostitute's schizophrenia, transfer CPF-OA to CPF-SA instead.

A quick way of resolving this schizophrenia is to transfer your CPF-OA to your CPF-SA. This is telling the prostitute that she is not expecting to do any housework so long as she services you well. In return, you end up with the best possible risk-less service possible in the Singapore economy at 4%.

The best time to perform the transfer is when you are young. You have more years to enjoy the 4% compounding. After all, the best times of life to get any serious bonking done is when you are much younger.

I maxed out my CPF-SA in my twenties. From the day I reached minimum sum, subsequent increases in the minimum sum has always been adequately met with the 4% returns. Even now, as my CPF-OA is almost totally exhausted maying off my home mortgage, I received over $10,000 in 2018 from interest alone. At my current age of 44, it does not take very long before I can reap the rewards of moving my CPF-OA to CPF-SA when I was in my 20s.

So there you have it, the most pornographic and obscene article you will ever read about the CPF program.

Just don't invite me to give a talk like this in public.






Friday, January 04, 2019

The Model Thinker #1 : The Wisdom Hierarchy

Image result for the model thinker

As we enter 2019, I will also be beginning a new series covering one really good book that I would like to digest very slowly. After some deliberation, I have decided to cover Scott Page's The Model Thinker  as it suits the level of an intermediate to advanced reader.

Naturally throughout the series, I will be personally trying to show-horn to adapt the ideas in this book to my personal style of investing. I hope that readers will find this new column useful and information.

Today I will be covering the Hierarchy of Wisdom. It is a great way for a novice to start thinking about models and how they can clarify things for us.

The Hierarchy of Wisdom consists of four ways in which we can explain the phenomena around us :

a) Data

Anything that lacks meaning can be considered data. Data can be a timestamp or a string of 1s and 0s. For a dividends investor, a 2 cts dividend payable on 2 December can be considered data since it is a timestamp of when money is to be credited into the shareholder's account.

On its own, a dividend does not have a lot of meaning.

b) Information

When data gets partitioned into categories, it becomes information. For example, when we sum up the dividend payments of a stock over a year, we get the total amount of dividends received in 2018. This becomes information. The information can be enriched further when we divide the annual dividends by the market price of a share and obtain the dividend yield of a company stock.

c) Knowledge

Knowledge is justified true belief.

When information can be correlated with each other or tied up with causal or logical relationships, we start to attain knowledge. If after a backtest, we find that picking dividend stocks tends to outperform the market index, this appears within the realm of knowledge. Deeper knowledge is gained when we figure out that picking for sustainable dividends from a high free cash flow can result in further outperformance.

Most financial providers show a student how to get to this level.

d) Wisdom

Wisdom is the highest level that can be attained. Wisdom is the ability to apply the right mental models to the proper situation. If you have his idea that higher dividends tend to outperform and have the tests to show that it works for equities, the question you need to ask yourself is whether the same principle can apply to REITs ( So far the backtests say yes. )

Problem arises is that you have many situations that use a regular cash-flow to tempt you into investing in something new. If you apply this model to bonds or crowdfunding projects, you will end up taking too much credit risk. Worse, there are now platforms that allow you to convert cryptocurrencies into a regular payment scheme.

Wisdom is to take a step back and perhaps understand that a better model to value these assets lies elsewhere.

e) I hope to be wiser in 2019

Some question in finance cannot be sufficiently answered with knowledge. One example is the question of whether a crash is imminent for the local stock-market. Academics have tried in vain to develop models to detect an inflection point in the stock-market but the best ones are, at best, accurate to about 75%.

I can't claim to teach wisdom to my class, but I adopt three models to provide an assurance that perhaps our markets are not in such a bad shape after all :

  • I examine the trending differences between the earnings yields of the SGX versus bond yields.
  • Next I look at CAPE Shiller PE of our markets to see if markets are overpriced.
  • Finally I look at the Sothebys stock and see whether it is experiencing a peak.  

The markets seem fine so far based on these measures.

So I don't wait for wisdom, I just stay invested and do not have a war chest for now.



Tuesday, January 01, 2019

Watch-outs and predictions in 2019



Image result for happy new year

One advantage of being a Christmas baby is that I get to reflect on the year on Christmas Day so it is possible for me to talk exclusively about the future on New Year's Day.

Instead of resolutions and goals, here are some cautionary statements entering the new year.

a) Don't talk about setting goals until you reviewed the ones you set last year

This insight came to me in a group chat with friends last week.

One guy was asking about what goals would we set for 2019 to which I then asked him what about the goals he set last year. The resulting silence was shocking and on hindsight, quite hilarious. You can almost hear a pin drop on Whatsapp.

It's a classic mistake. Even in business, there is no point setting KPIs if there is no way to monitor it throughout the year. A review will often flag many goals unmet and priorities altered. I think it's fine to miss a few goals but at least figure out why they were not met.

So I think it's nice to have goals and resolutions for 2019, but I think the first goal is to review your goals at end-2019. Otherwise no one will know how well you do. Least of all yourself.

b) Make sure your hardware matches your software

Another insight came from almost killing my Oneplus 2 phone. If a cheap 4 year old phone can run Android Pie, what is more important to a person ? Hardware or software ?

I know folks who spend quite a bit on sound systems. This is a hobby with no real upper limit. One of those financial crippling hobbies for rich people.

Surprisingly, when I ask audio aficionados about choice of music, I don't seem to get the same enthusiasm. The truth is that some people buy systems costing $15k and above just to play elevator music. The same way you pay $1,300 to buy a Pixel 3 XL just to Whatsapp your friends.

Of recent note, I have a new explanation for this behaviour.

This is a mismatch between a person's financial and cultural capital.

As your financial capital goes up, you have the money to pick up the best gear and gadgets that money can buy. However, your cultural capital may still be stuck in the 1980s. Developing cultural capital takes time and requires a lot of curation. You also need a relatively high openness to new experiences. For folks with a low openness to new experience, it becomes harder to pick up a new artist or musical genre.

Of course, it might actually be cool if you buy a top of the line AK Audio system just to play Rick Astley songs, but the same amount of time and effort can be used to cultivate a new kind of taste in music.

If you want to appreciate new forms of entertainment, where do you start ? I think 1843 magazine, the cultural variant of the Economist has suggestions on what kind of music to listen to.

As a 40-something, I am pretty closed minded to new things myself, but I am still trying.

c) Accept truths only if they are useful.

Not all truths are useful.

Of late, some gurus have been claiming that investing is more of an art than a science. As much as I hate this, I have to agree that this statement is largely true because even my own FK Stack suggests a personal and discretionary dimension to financial decision making.

Saying that investing is an art, even if true, is not useful. Worse, if you teach investing, it can be used as a cop-out to just explain why the student underperforms the market.

It is also a grievous insult to social science and humanities majors who study the Arts towards becoming useful to society.

I propose a better response against folks who claim that investing is an art.

Even in the humanities and the  social sciences, models exists to at least explain something. These explanations may not be complete so they need to be supplemented with other mental models.

Next time someone says that investing is an art, ask them for a number of models that may partially explain the phenomenon of out performance.

Suppose a "guru" says that finding an investment moat is an art. Don't let him off. Ask him for about 2-3 examples of companies with an investment moats and drill him until he can generalise a few traits about these companies. If you confront me, I might say that licensing barriers can constitute a moat. Alternatively, you can look for networking effects.

A real "guru" might be able to come up with a checklist to at least flag the moats a company can possibly have.

d) Finally, predictions for 2019

I now realise that by the end of the month I will be completely off with my market predictions, but it would still be fun to at least make a guess at this point of time.

I predict that markets will continue to trend downwards for 1H2019 and recover in 2H2019, with a final STI range from 3100 to 3200 at end-2019. However, investing in emerging markets should do well throughout the year as they are already very cheap.

Also Cryptocurrencies will either stay flat or do worse in 2019. They remain solutions looking for a problem and you might want to give the next ICO a miss.