Friday Funny – Investing Vs. Gambling

This will never work (Vol 1, No.8)


There is a fine line between investing and gambling. For most retail investors, buying a stock is really no different than betting on your favourite number on the roulette table.

Things in common – math and statistics

What they both have in common is math. There’s a bunch of statistics that we can fall back on. We know what the long term stock market returns are like and we also know  what are the odds of the winning at baccarat. But what we don’t know is what will happen in the short term?

Key difference – Positive versus negative long term returns

What’s different is that in investing, the longer you stay invested in the markets, the more likely you are to see positive returns. On the contrary, as the odds favour the house, the longer you gamble, the more likely you are to see negative returns.

For instance, we know that from 1987-2017, out of 7,659 trading days, the Hang Seng Index had risen 52% of the time. We also know that the average 1.08% gain on those Up days outweighed the average down day’s -1.072% correction and the resultant average +0.046% daily return compounded nicely over time.

chart (64)

Conversely, in the gambling sphere, with rolling chip win percentage around 2.7-3.3%, compounding works in the favour of the casinos.

What’s funny about that?

Hang on a minute here, isn’t the title Friday Funny? This lecture doesn’t sound funny at all.

Well, the whole point of the above is just for me to introduce this YouTube clip. It’s from three years ago but doesn’t feel dated at all. This goes out to my Singaporean friends.

Update on “Mean Reversion”

On a more serious note, the stock market has continued to misbehave.

In mid-February, in the wake of the market sell-off that erased 10.6% over 13 days, we ran some numbers and basically gave a non-answer to whether stock markets were due for a bounce (see link here).

Two months and 48 trading days have passed and we are pretty much back where we started. The Hang Seng Index bounced back from 29,540 to as high as 31,601 (up 7.0%) and has since fallen back to 30,008 (5% retracement).

Does this mean that we have “mean reverted”? Is the volatility over?

Up Days to Down Days – The ratio of Up Days to Down Days has gone from 19:11 to 45:33. As up days still make up 58% of 2018’s trading days, this is still above the long run average of 52% and suggest more rainy days may lie ahead.

chart (65)

Daily Percentage Change – Thankfully, the amplitude of the down days have moderated somewhat. Instead of falling an average of -1.277% (over the first 11 down days), the average down day has now moderated to -1.176% (over 33 down days). At the same time, the average gain on the up days have picked up from +0.673% (over the first 19 Up days) to 0.883% (over 45 days).

chart (66)

My suspicion is that our earlier advice to develop a hobby as a coping mechanism to deal with higher volatility still stands. If not, there’s plenty of funny stuff on YouTube.

Happy Friday!




The timeframe to be “Mean”…

Mean (noun) – The value obtained by dividing the sum of several quantities by their number; an average.          Oxford Dictionary

Bet you thought I was talking about “mean” the adjective. No, I wouldn’t want to spoil the joyful mood around Chinese New Year with talk of being unkind, spiteful or unfair.

Instead, this post is about “Mean Reversion”.

At this moment, I am actually lying by the pool somewhere in the tropics and away from the cold. This piece was actually pre-written a week ago, so the data and the topic could seem dated but anyways you got to make some trade offs here and there.

Stock markets due for a bounce?

Time frame #1 – The last 13 days suggest YES

So cast your mind back to Monday February 12. After a torrid run in January, the Hang Seng Index has fallen in 10 out of the last 13 trading days. Over this period, the HSI has declined from 32,959 to 29,540. The cumulative 10.6% decline has officially put us into correction territory and also pushed 2018 into negative territory at -1.5%.

chart (13)

So, are we due for a bounce? Well, if you only consider the last 13 days, then YES. As we have previously written, in the long run, the number of up and down days are almost 50/50 (specifically from 1987-2017, the ratio has been 52:48 – see previous post here).

Timeframe #2 – 2018-to-date suggest NO

But not so fast. If we just stretch the timeframe back slightly to the start of 2018, we can see that despite the market having fallen in 10 of the last 13 days, the number of UP days still outnumber the number of DOWN days 19:11. Yikes.

Timeframe #3 – 2017-to-now suggest NO

But what if we go back even further? I’m afraid that still doesn’t do it for us. You see, 2017 was a pretty good year for stocks and the Hang Seng Index’s 35% gain makes the Hong Kong market one of the best performing one last year.

chart (14)

In 2017, we had 142 UP days and only 102 DOWN days. The 58:42 ratio was only second to 2006’s 60:40 ratio and only equaled by the market’s performance in 1993. Prior to last year, the HSI’s performance had been pretty middling. From 2012 to 2016, there were 623 UP days and 606 DOWN Days for a 51:49 ratio. So, mean reversion would suggest that there ought to be more down days ahead.

Another Frame of Reference Suggest YES

So is there any good news? Yes. The clue lies in the details from Timeframe 1 and Timeframe 2. Why is it that despite the number of UP days outnumbering DOWN days by 19-to-11 is the market down 1.5%? The answer is simple. The magnitude of the correction has been furious. It’s the amplitude and not the frequency.

For the 11 DOWN days in 2018, the average daily correction has been -1.277%. This is more than double the -0.51% average daily move during 2017’s 104 DOWN days.

In the past, when we have looked at how the average UP days have compared to the average DOWN day, the average UP move of 1.08% has exceeded the average DOWN move of 1.07% marginally. The combination of a 1.01x UP/DOWN percentage ratio, coupled with a 52:48 UP/DOWN day count plus the magic of compounding have helped the HSI gain 10x over the past 30 years. chart (15)

What is striking for 2018 is how the sharp correction of the past two weeks have thrown off the UP/DOWN percentage ratio. In 2018, the average UP day percentage change has only been 0.673%. Compared to the -1.277% average correction for the 11 Down days, the current UP/DOWN percentage ratio of 0.53x is too negative. If this mean were to revert, the UP day gains either have to get a lot bigger or the DOWN day corrections have to become a lot milder.

Get a hobby, stress level looks set to increase

Before we finish, there is one more important detail from the previous chart. 2017 was a good year. Not only did the HSI gain 35% but this was done with very little excitement. The market just plodded steadily upwards. The average gain on up day was only 0.595% while the average down day was only -0.511%.

The last time things were this calm was in 2005. While it is true that the 2005 calm preceded another 1-2 years of gains, it is clear that the market rarely stay calm for long. If the past cycle holds, we better pick up some hobbies to distract us from the wild market swings that are likely on the way.

chart (16)

With that, I think I have managed to achieve the Holy Grail of market commentators, that is write a piece that will be proven right whether the market goes up or down. Or in other words, told you a lot of stuff without actually telling you anything.




From 50/50 to 59/41 – Return of the animal spirits

The Hang Seng Index has risen 25% in 2017 and is now within 3.7% of the 2015 peak. The market had been strong but how much of an outlier is it? When we look at daily performance, we find the HSI had risen 59% of the time with an average gain/loss ratio of 1.29x, putting it in similar outlier territory as 2015 and 2007.

Until very recently, Hong Kong’s 2017 stock market rally had been one of the most stealthy bull market in recent memory. As of  1 August, the Hang Seng Index had risen 25% and at 27,540, the HSI is only 3.7% below the 2015 peak of 28,588 and 13.8% below the all-time high of 31,958 that was reached on 30 October 2007.

Screen Shot 2017-08-03 at 12.42.26 am

3.7% away from 2015 peak, will the good times last?

As we approach the 2015 大時代 peak, we get the inevitable question about bubbles and whether we are due for a correction.

From a historical perspective, we can certainly appreciate where the concern is coming. In 2015, the Hang Seng Index had a blazing start, rising 21% by 28 April. Had the pace continued for the full year, it would have annualised to 63%. But the good times did not last. After peaking on April 28, the market corrected 23% the rest of the year and reversed the previous gains to a full year loss of 7.4%. Ouch.

Screen Shot 2017-08-03 at 9.37.58 am.png

So, what are we to make of this year’s (2017) strong performance? Is it a repeat of 2015’s roller coaster ride or is the current rally likely to prove more sustainable?

Until today, my gut feel was that it may have more room to run. Why? First, I think the market’s animal spirits feeds on recent peaks. Until 2015’s 28,588 peak is surpassed, I suspect Mr. and Mrs Wong are still sceptical. From a behavioural perspective, others have commented that one sign of bubble is when taxi drivers start to give out stock tips and people quit their day jobs to day-trade.

I read somewhere that in the US market, the number of days that the market goes up and the times that the market goes down is roughly 50/50. Although the odds are roughly even, the reason why the markets have tended to go up is because average gains have outweigh average losses and the compound effect leads to long-term cumulative gains.

3,922 up days and 3,633 down days since 1986

I wanted to see if the same held true for the local stock market. Looking at the Hang Seng Index’s performance since 1986, although the market had risen 9.7x, the number of up days versus the number of down days is almost 50/50. Specifically, there were 3,922 up days and 3,633 down days. This is a ratio of 51.9%/48.1%. If one were to consider the average daily movements, the average up day gain was 1.09% while the average down day loss was 1.08%, so again very  close to 50/50.

Screen Shot 2017-08-03 at 9.34.03 am.png

2017 has been an outlier – 59.4% Up Days and Gain/Loss ratio of 1.29

What about 2017? Well, in the 143 trading days this year, there has been 85 up days and 58 down days. This is a ratio of 59.4/40.6. Furthermore, when we consider the average daily movement, although the average gain on up days is only 0.57%, when this is compared against an average loss of 0.44% on down days, the gain/loss ratio of 1.29 is much higher the past five year’s average of 1.01.

Screen Shot 2017-08-03 at 1.05.08 am

Current rally as much of an outlier as 2015 and 2007

Clearly, the 2017 rally had been strong but how does it compare to the previous “bubbles”. In the more recent memory, there was the 2015 大時代 rally. Through the first 78 trading days of 2015, the index rose 21%. Over those 78 trading days, there were 49 up days and only 29 down days. While 2015’s up days ratio of 63% is 4pp higher than 2017’s 59%, the average gain/loss ratio was milder at 1.19x (0.77% average gain versus -0.64% average loss).

Screen Shot 2017-08-03 at 1.06.24 am

If one were to look further back, we saw an even more euphoric rally in 2007. Driven by hopes of a “through-train’ of China domestic investments, the Hang Seng Index had rallied by 60.7% through 30 October 2007 and reached an all-time intra-day high of 31,958. During this period, over the 204 trading days, the Hang Seng Index rose on 115 days and fell on 89 days. The Up days to Down days ratio was 56/44. Although this was milder than 2015’s 63/37 up/down days ratio, the longer duration of the cycle couple with an average gain/loss ratio of 1.20x helped to push the market up by 60%.

2017 less hot than 2005 and 2007

So if one were to consider the current market condition, we can see that 2017’s up days ratio of 59.4% is higher than 2007’s 56% but lower than 2015’s 63%. But from a daily percentage change ratio, 2017’s 1.29x ratio is higher than 2015’s 1.19x ad 2007’s 1.20x.

Screen Shot 2017-08-03 at 1.08.35 am

In the long run, reversion to the mean is likely but timing is anyone’s guess

In the long run, a reversion to the man is probably inevitable and the up day to down day ratio should probably trend back towards 50/50. But as one can see in the table above, that up days ratio remained elevated in 2006 (59%) and 2007 (56%) and only reverted back to 47% in 2008.

Be that as it may, given the combination of higher than expect win streak and higher than normal win ratio, it might be better to simply ride the current wave than to initiate new buy positions at this stage. Caveat emptor.