Sunday, 19 February 2017

The Telco Landscape in Singapore

The telco regulator in Singapore, Infocomm Media Development Authority (IMDA), regularly publishes a list of facts & figures about the telco industry. It is interesting to look at the figures occasionally to understand the trends affecting telcos and review whether telcos worth investing.

Mobile Services

To investors of the 3 local telcos, the most important figure must be the no. of mobile subscriptions. Fig. 1 below shows the no. of subscribers according to type of network (3G/4G) and payment mode (pre-paid/ post-paid) for 2016 up to Nov.

Fig. 1: Mobile Subscriptions

From the figure, it should come as no surprise that 4G subscriptions are rising as more subscribers switch from 3G to 4G. However, what is surprising is that the 3G pre-paid segment (red line in Fig. 1 above) is fairly resilient, dropping only by 3.4% over the 11 months of 2016 even as the 3G post-paid segment shrank by 22.1%. The 3G pre-paid segment remains the second largest market segment, commanding a market share of 32% of the mobile services market.

Why do 3G pre-paid subscribers choose not to upgrade to the 4G pre-paid network? To mobile subscribers, the main difference between the 3G and 4G networks is faster data throughput on the 4G network. Thus, subscribers who use mobile data will be keen to upgrade to the new network. However, for subscribers who use only voice and SMS but no data, there is very little difference between the 3G and 4G networks. Given that pre-paid subscriptions do not come with much data (you need to buy a data add-on), there are little incentives for 3G pre-paid subscribers to move on to the 4G network.

Internet Broadband

Besides mobile services, telcos also sell internet broadband services, in the form of fibre, cable and xDSL broadband. Fig. 2 below shows the no. of broadband subscribers for the 11 months of 2016.

Fig. 2: Broadband Subscriptions

Again, it should come as no surprise that fibre broadband subscriptions are rising as they offer faster speeds of 300Mbps or more. This growth is at the expense of cable and xDSL broadband, which Starhub and Singtel operate respectively. Thus, Starhub and Singtel should be wary of declining revenue from their respective cable/ xDSL broadband business. Although both will gain from increased fibre broadband business, the fibre broadband segment is a competitive one with many operators such as the 3 telcos and MyRepublic. Telcos have found it necessary to bundle mobile broadband and home digital lines to improve their offerings.

Finally, there is 1 particular chart that all wired broadband operators should worry about, which is that the no. of residential wired broadband subscriptions have peaked in 3Q 2015 (see Fig. 3 below)!

Fig. 3: Residential Wired Broadband Subscriptions

Where did broadband subscribers go to for their internet consumption? Fig. 4 below provides the answer.

Fig. 4: Wireless Broadband Subscriptions

Fig. 4 shows that even as broadband subscribers begin to terminate their wired broadband subscriptions, they have continued to subscribe to mobile broadband. This reflects changing internet consumption habits of consumers, who access internet and watch online videos on the go instead of being tied to accessing them at home. 

The penetration rate of mobile broadband as at Nov 2016 is 196.7%, even higher than that of mobile services at 149.2%! There are more mobile broadband subscriptions than mobile service subscriptions! Again, this reflects more people having more mobile devices such as phones, tablets, smartwatches, etc.

Entry Points for 4th Telco

A key reason for the analysis above is to understand whether there are any market segments for TPG, the fourth telco, to exploit and enter the Singapore market in a big way. When MyRepublic made its way into the fibre broadband market, it did so at a time when the fibre broadband market was still nascent and the telcos were reluctant to reduce prices so as not to cannibalise their existing broadband businesses. This provided MyRepublic a chance to slash prices aggressively, gain widespread publicity and lead the competition. 

Looking at the figures above, the 3G pre-paid market segment is a potential market that TPG could exploit. It is 32% of the mobile services market. Currently, none of the existing telcos could entice 3G pre-paid subscribers to move to 4G in a big way. If TPG could offer a compelling reason to 3G pre-paid subscribers, it could capture large market share from the 3 existing telcos and entrench itself in the mobile services market.

In the fibre broadband market, competition is already very intense. While TPG will enter this market and provide fibre broadband services, I do not expect any innovative offerings. 

In the mobile broadband market, the penetration rate is already very high at 196.7%, which is equivalent to each person having 2 mobile broadband subscriptions. Having said the above, the next big bang will be the Internet of Things, whereby everything is connected to and streaming data to the internet. When this happens, the market for mobile broadband will expand significantly and all telcos, including TPG, will stand to gain.

P.S. I am vested in M1 and Singtel.

See related blog posts:

Sunday, 12 February 2017

Can Telcos Stop the Decline in Profitability?

All 3 telcos have reported their financial results for the quarter ending Dec 2016. Both M1 and Starhub reported a significant drop in net profit of 27.1% and 33.2% respectively while Singtel reported a slight 1.8% rise in net profit. Starhub even dropped a bombshell by announcing that dividends for 2017 will be cut by 20% to 16 cents. Can telcos stop the decline in profits, or will profits continue to decline? As M1 is closest to a pure mobile telco company, this analysis is carried out using M1's results.

Fig. 1 below shows the overall service revenue and costs excluding handset sales, which are very volatile. As shown below, the service revenue is fairly stable, hovering around $200M per quarter in FY2015 and FY2016. Service costs, however, are on the rise, especially in the last 2 quarters, leading to a decline in quarterly profits.

Fig. 1: M1's Service Revenue & Cost

Service Revenue

Fig. 2 below shows the breakdown of the service revenue. The bulk of the revenue comes from mobile services, which constitute 79% of the total service revenue. The rest of the service revenue comes from IDD calls and fixed services (i.e. fibre broadband).

Fig. 2: Breakdown of Service Revenue

As shown in the figure above, mobile services revenue is on the decline, which explains why the telco profits are declining. Some of the reasons for the decline are discussed in Impact of SIM-Only Plans on Telcos and Impact of Data Upsize Plans on Telcos. Since this business segment has the largest impact on telco profitability, I will come back again to discuss the trends affecting this segment.

IDD revenue is on the decline, which is not surprising given the popularity of over-the-top (OTT) services such as Whatsapp, which allows any Whatsapp user to make a phone call to fellow Whatsapp users anywhere in the world using WiFi or mobile data instead of voice. Very likely, this downward trend in IDD revenue will continue.

On the other hand, revenue from fixed services is growing strongly as more people subscribe to fibre broadband. In the short term, this segment is still expected to grow, but perhaps not as strongly as before as it faces many competition from other telcos as well as broadband service providers like MyRepublic. Already, M1 has to bundle mobile broadband and home digital voice to improve its offering among the competition.

Service Costs

Fig. 3 below shows some of M1's largest cost components necessary to earn the service revenue discussed above.

Fig. 3: M1 Service Costs (Extracts)

As shown in Fig. 3, depreciation and amortisation costs, which is the largest cost component, is on the rise, as M1 continues to invest in building its network. In fact, in the latest financial results, M1 reported that the estimated capital expenditure (capex) for FY2017 is around $170M. For FY2016, the capex was $141.2M, excluding a spectrum rights payment of $64.1M. Depreciation is definitely on the rise moving forward. 

The next largest cost component, staff cost, is fairly stable. In my opinion, this cost component is unlikely to rise, because TPG, the fourth telco, would likely not have any physical stores and would rely on third-party dealers to sell its services. The 3 current telcos, which have their own physical stores and sales staff, would be under pressure to contain staff cost.

The third largest cost component, facilities expenses, is also on the rise. I am not sure if facilities refer to physical stores or telco network facilities. If it is referring to physical stores, this cost will be contained for the same reason as discussed above for staff cost. However, if it is referring to telco network facilities, this cost will be on the rise in tandem with capex. To be conservative, I assume that it is referring to telco network facilities and will continue to rise.

The last major cost component, fixed services cost, is rising as well. However, this is of no concern, as the rise is in tandem with the rising revenue from fixed services shown in Fig. 2. Deducting the cost of $11.2M from the revenue of $27.2M, this segment earned a gross profit (excluding all sales-related costs like staff costs, rental leases, etc.) of $16.0M for 4Q2016. This is a rise from $10.8M in 1Q2015.

Revenue Drivers

Based on the discussion so far, service costs are likely to continue rising due to expansion of the network infrastructure. Thus, if M1 is unable to halt the decline in service revenue in its main mobile services segment, it is likely face declining profits and dividends, even before the fourth telco opens for business.

2 of the factors affecting the mobile services revenue have been discussed in previous blog posts, namely, Impact of SIM-Only Plans on Telcos and Impact of Data Upsize Plans on Telcos. Both these factors are negative on revenue and profitability. The impact of SIM-only plans will be spread out over 2 years, as subscribers can only switch from regular plans to SIM-only plans when their contracts expire, and the typical contract period is 2 years. On the other hand, the impact of data upsize plans is fairly immediate, as subscribers need not wait for their contracts to expire before subscribing to the upsize plans. Having said that, when comparing results on a year-on-year (YOY) basis, it will take 1 more year for the effect to fully disappear from the financial statements.

Although there are headwinds from SIM-only plans and data upsize plans, there are, nevertheless, silver linings. As previously discussed in Impact of Data Upsize Plans on Telcos, the larger data allowance is encouraging subscribers to use more data, as shown in Fig. 4 below (note: M1 changed the way it measures the percentage of subscribers who exceed their original data allowance in FY2016. The figures for 2015 and 2016 are not directly comparable).

Fig. 4: M1's Subscriber Data Usage

Perhaps as a result of the introduction of SIM-only plans, the number of post-paid subscribers has also jumped. The blue line in Fig. 5 below shows the YOY growth in number of post-paid subscribers, which jumped from 1.8% in 2Q2015 to 3.0% in 3Q2015, the quarter when M1 introduced SIM-only plans. The growth has continued to rise and recently stablised at around 4.5%. This rise in post-paid subscribers will counteract the fall in revenue due to conversion from regular plans to SIM-only plans by existing subscribers and data upsize plans.

Fig. 5: Trends in Post-Paid Customer Revenue

Also on the rise is the percentage of post-paid subscribers on tiered data plans (see red line in Fig. 5 above). Prior to the introduction of tiered data plans (and smartphones), some subscribers had very large data allowances under their very old mobile subscription plans. Tiered data plans cap the mobile data allowance according to the monthly price plans and subscribers have to pay more for having more data.

In conclusion, there are both headwinds and tailwinds for the mobile services operations. 

Finally, please note that although all 3 telcos have mobile services operations, Starhub has other business segments and Singtel has overseas operations and investments. 

P.S. I am vested in M1 and Singtel.

See related blog posts:

Sunday, 5 February 2017

Impact of Data Upsize Plans on Telcos

Competition among the 3 local telcos used to be fairly stale, until things started to heat up in Mar 2016 when all 3 telcos launched data upsize plans as a way to pre-empt competition from a potential fourth telco. The upsize plans provide subscribers with more data with the payment of an extra monthly fee. How does that impact the revenue and profitability of telcos? As M1 is closest to a pure mobile telco company, with 79% of its service revenue derived from mobile telco services, this analysis is carried out using M1's results.

Data X2

The table below shows the monthly cost and data allowance of M1's regular plans without data upsize and with Data X2 upsize (Note: each telco calls its plans differently and has different data sizes. I am using Singtel's naming convention for ease of reference). M1's Data X2 upsize costs an additional $5.90 monthly for all regular plans.

Plan Lite Lite+ Reg Reg+ Max Max+
Monthly Cost $28.00 $42.00 $62.00 $82.00 $102.00 $228.00
Original Data (GB) 0.3 3 4 5 7 13
Data X2 Cost
$47.90 $67.90 $87.90 $107.90 $233.90
Data X2 (GB)
5 7 9 13 25

Prior to the launch of data upsize plans, M1 charged $10.70/GB for excess data usage beyond the data allowance. Thus, a subscriber who previously had to pay $10.70/GB for excess data usage now only has to pay $5.90 to upsize his data allowance. He saves at least $4.80 per month on excess data usage. This has a direct impact on M1's revenue and gross profit, since the network infrastructure is already set up and the marginal cost of providing the additional data is likely to be small.

Using M1's financial results for 4Q2015, which is just before the 3 telcos launched their data upsize plans in Mar 2016, the percentage of subscribers who exceeded their data allowance was 21%. I assume conservatively that these subscribers exceed their data allowance by not more than 1GB, because those who exceed their data allowance by between 1-2GB would have to pay $21.40 more in excess data usage and are better off subscribing to the next tier regular plan, which costs $20 more only. Thus, M1 potentially loses $4.80 per month for every subscriber who chooses to upsize their data allowance. As at 4Q2015, the number of post-paid subscribers was 1.195M. The number of subscribers who exceeded their data allowance was 251K (21% of 1.195M). The potential revenue loss is 251K x $4.80 x 12 months or $14.5M. This works out to be 2.4% of FY2015's revenue of $591M. It is also equivalent to 6.6% of FY2015's pre-tax profit of $218.4M.

Unlike the SIM-only plans, which subscribers could only switch when their 2-year contracts expire, subscribers could choose to upsize their data allowance at any time. Thus, while the impact of SIM-only plans are spread out over 2 years, the impact from data upsize plans is fairly immediate. 

Fig. 1 below shows the year-on-year changes in M1's revenue and post-paid ARPU, with timeline of the launches of the SIM-only plans and Data X2/X3/X4 plans superimposed on the chart. To understand the impact of SIM-only plans, please refer to Impact of SIM-Only Plans on Telcos.

Changes In M1's Revenue & ARPU

From Fig. 1, M1's revenue and ARPU have actually started falling in 2Q2015, but the decline accelerated from 3Q2015 onwards, which was when M1 launched the SIM-only plans. Things got worse in 1Q2016, which coincided with the launch of data upsize plans. 

Given that the impact of data upsize plans is fairly immediate, I am unable to explain why M1's revenue continued to fall at a rapid pace after 2Q2016. The reasons M1 gave for the decline in net profit were lower IDD and roaming revenue, higher handset subsidy and higher depreciation and amortisation. Thus, there are other factors at play besides SIM-only plans and data upsize plans.

Data X3/X4

As if Data X2 plans are not good (or bad) enough, Singtel launched Data X3 plans in Sep 2016, just after TPG, MyRepublic and airYotta submitted bids to be the fourth telco. M1 followed suit in Nov 2016 with not just Data X3 but also Data X4 plans!

As M1 did not disclose the number of subscribers who exceed their Data X2 allowance, it is difficult to assess what is the impact to revenue and profitability of Data X3/X4 plans. My personal opinion is the number of such subscribers is likely to be small and thus, the impact on revenue and profitability is likely to be small as well. Data X3/X4 plans are likely to be marketing gimmicks.

What's Next

Data upsize plans may be bad for telcos, but there is a silver lining. The higher data allowance of the upsize plans are enticing some subscribers to use more data. Fig. 2 below shows the average data usage and percentage of subscribers who exceed their original data allowance.

Fig. 2: M1's Subscriber Data Usage

As shown in Fig. 2, there is an uptick in average data usage after 2Q2016, which is after the launch of data upsize plans in Mar 2016. The percentage of subscribers who exceed their original data allowance is also on the rise (note: M1 changed the way it measures this metric in FY2016. The figures for 2015 and 2016 are not directly comparable). This will eventually translate to increased revenue and profitability as subscribers who now exceed their original data allowance and upsize have to pay at least $5.90 extra per month.


Data upsize plans have a direct and immediate negative impact on telco's revenue and profitability. However, they are also enticing subscribers to use more data, which will gradually translate to higher revenue and profitability (assuming all other factors remain constant).

Finally, please note that although all 3 telcos have mobile telco operations, Starhub has other business segments and Singtel has overseas operations and investments. 

P.S. I am vested in M1 and Singtel.

See related blog posts:

Monday, 30 January 2017

Impact of SIM-Only Plans on Telcos

I recently switched my telco subscription plan from the regular ones to SIM-only plans. How do these SIM-only plans impact the revenue and profitability of the 3 local telcos?

Before we begin, let me explain what are SIM-only plans. As the name implies, SIM-only plans only provide voice, SMS and data but do not come with subsidised handphones. The tables below show the amount of voice, SMS and data for regular and SIM-only plans for M1, the telco which I subscribe to. The last table shows the difference in the amount of voice, SMS and data between the 2 types of plans. As you can see, SIM-only plans have generally the same amount of voice but less SMS and more data than regular plans. As people seldom SMS but rely more on data nowadays, I would argue that SIM-only plans provide better value than regular plans at lower costs!

Regular Plans

Plan Lite Lite+ Reg Reg+ Max Max+
Monthly Cost 28 42 62 82 102 228
Voice (mins) 100 200 300 400 800 Unlimited
SMS/MMS 500 1000 1200 1500 2000 5000
Data (GB) 0.3 3 4 5 7 13

SIM-Only Plans

Plan MS+ 15 MS+ 20 MS+ 30 MS+ 45 MS+ 75 MS+ 125
Monthly Cost 15 20 30 45 75 125
Voice (mins) 100 150 300 400 800 Unlimited
SMS/MMS 600 800 1000 1200 2000 Unlimited
Data (GB) 1 4 6 8 13 20


Monthly Cost -13 -22 -32 -37 -27 -103
Voice (mins) 0 -50 0 0 0 0
SMS/MMS 100 -200 -200 -300 0 Unlimited
Data (GB) 0.7 1 2 3 6 7

What is the impact of SIM-only plans on telco's revenues? We have to assume a particular handset to illustrate the impact. Let us use Samsung S7 as the handset. If you purchase it under the regular plans, the price you pay ranges from $608 to $0, depending on which plan you choose. This sale of handset contributes to the overall revenue of the telco. However, for SIM-only plans, there is no sale of handset, so handset sale no longer contributes to the overall revenue. Not only that, the service revenue is also lower. The table below shows the revenue impact of SIM-only plans. Using Lite/ mySIM+ (MS+) 15 plans as an example, the total revenue over a 24-month period is $1280 under the regular plan, but only $360 under the SIM-only plan. The decline in revenue is $920, or 72% of the revenue under the regular plan! As we move towards the more expensive plans, the percentage decline becomes lower. Thus, we can expect telco revenue to drop, depending on how many people switch to SIM-only plans.

Lite Lite+ Reg Reg+ Max Max+

MS+ 15 MS+ 20 MS+ 30 MS+ 45 MS+ 75 MS+ 125
Handset Price

Regular 608 458 128 58 0 0
SIM-Only 898 898 898 898 898 898
Total Revenue

Regular 1280 1466 1616 2026 2448 5472
SIM-Only 360 480 720 1080 1800 3000
Difference -920 -986 -896 -946 -648 -2472
% Difference -72% -67% -55% -47% -26% -45%

For M1, the expected decline in revenue shows up in their latest financial statement for FY2016 released last week. Fig. 1 below shows that handset sales dropped by 23.7% in FY2016 and mobile telco revenue dropped by 4.2% despite the total no. of mobile customers rising by 4.7%! Likewise, Fig. 2 below shows that the Average Revenue Per User (ARPU) dropped by 6.0%.

Fig. 1: M1's Revenue

Fig. 2: M1's ARPU

Having said the above, please note that the reasons M1 gave for the decline in net profit were lower IDD and roaming revenue, higher handset subsidy and higher depreciation and amortisation. So, perhaps not many people switched over to SIM-only plans.

Although revenue is expected to drop with SIM-only plans, if the cost of sales drop correspondingly, the impact to profit would not be significant. Here, I will attempt to estimate the profit impact from SIM-only plans. A few assumptions need to be made. Firstly, I will assume that M1 is a pure mobile telco company, even though it also has international call services and fixed services, which together constitute 21% of its service revenue in FY2016.

In FY2016, M1 reported gross profit of $180.0M on total revenue of $1,060.9M, giving it a profit margin of 17.0% (Based on FY2014 financial results, which is before M1's launch of SIM-only plans in Jul 2015, the gross profit margin was 20.4%. We will use the more conservative and latest figure of 17.0% in this analysis). Since this profit margin includes both the sale of handsets and mobile telco services, it is assumed to be the profit margin for the regular plans.

Overall Handset Service
Revenue 1060.9 255.4 805.5
Expenses 880.9 343.9 537.0
Gross Profit 180.0 -88.5 268.5
% Profit 17.0% -34.7% 33.3%

A breakdown of the revenue and expenses shows that handset sales contributed $255.4M to the overall revenue and $343.9M to the costs. In other words, M1 subsidised $88.5M for the sale of handsets to its regular plan customers. Removing the revenue and expenses from handset sales, the service revenue is $805.5M and the cost of service is $537.0M, giving it a gross profit of $268.5M or a profit margin of 33.3% on mobile telco services alone. This is assumed to be the profit margin for SIM-only plans, since they only provide mobile telco services.

Finally, I will further assume that the profit margins of 17.0% and 33.3% apply for all regular and SIM-only plans respectively, which is likely to be incorrect. I understand that a lot of assumptions have been made here, but these assumptions provide at least a starting point for analysing the impact of SIM-only plans on telco profitability. If you have better figures, please let me know.

Applying these profit margins to the revenue discussed above, the gross profit for the various plans are shown in the table below.

Lite Lite+ Reg Reg+ Max Max+

MS+ 15 MS+ 20 MS+ 30 MS+ 45 MS+ 75 MS+ 125
Gross Profit

Regular 217 249 274 344 415 928
SIM-Only 120 160 240 360 600 1000
Difference -97 -89 -34 16 185 72
% Difference -45% -36% -12% 5% 44% 8%

Using the Lite/ MS+ 15 plans as an example, the gross profit for the regular plan is $217 (17.0% x $1280) but only $120 (33.3% x $360) for the SIM-only plan. The difference is $97 or 45% of the gross profit of the regular plan. However, as we move towards the more expensive plans, this difference turns from a loss into a profit! The reason is pure mobile telco services has a higher profit margin. Thus, the more expensive the plan is, the more profit the telco makes!

The same analysis repeated for iPhone 7 (128GB) is shown below. Likewise, it shows that there is a significant drop in revenue but the impact to profit depends on which plan the subscriber chooses.

Lite Lite+ Reg Reg+ Max Max+

MS+ 15 MS+ 20 MS+ 30 MS+ 45 MS+ 75 MS+ 125
Handset Price

Regular 880 745 560 380 190 0
SIM-Only 1218 1218 1218 1218 1218 1218
Total Revenue

Regular 1552 1753 2048 2348 2638 5472
SIM-Only 360 480 720 1080 1800 3000
Difference -1192 -1273 -1328 -1268 -838 -2472
% Difference -77% -73% -65% -54% -32% -45%
Gross Profit

Regular 263 297 347 398 448 928
SIM-Only 120 160 240 360 600 1000
Difference -143 -137 -107 -38 152 72
% Difference -54% -46% -31% -10% 34% 8%

So, the next 2 big questions are: (1) how many people will choose to switch from regular plans to SIM-only plans, and (2) which plans are they on currently?

I do not have much insights on these 2 questions. On the first question, my opinion is that people who choose to switch include those who are cost-conscious, do not need an expensive phone and/or tired of upgrading phones every 2 years. The possible profiles of these groups of people would be the elderly and young children (if they are not already on the pre-paid plans). To cut a long story short, I do not think a lot of people will switch.

Another point to note is that M1 launched the SIM-only plans in Jul 2015. Since the typical contract period is 2 years, we are approximately 75% into the first renewal cycle. Nevertheless, it is also possible that some subscribers could have missed the SIM-only plans when they previously renewed and choose to make the switch in the next cycle.

On the second question, based on Fig. 2 above, the ARPU is $58, which is closest to the Reg plan that costs $62 per month. For this plan, the SIM-only equivalent will result in lower profit based on the analysis for Samsung S7 and iPhone 7 (128GB) above.

Hence, in conclusion, for each subscriber who chooses to switch, SIM-only plans will result in a significant drop in revenue. The impact to profit is also negative but smaller than the decline in revenue. However, the no. of subscribers who choose to switch to SIM-only plans is likely to be small.

P.S. I am vested in M1.

See related blog posts:

Sunday, 22 January 2017

A Prediction About Properties 13 Years Ago

Before I became a blogger in 2012, I contributed to the Straits Times (ST) Forum Page once in a blue moon. My first letter was written in Dec 2004 and discussed the issue of whether downgrading homes was a viable means of providing for retirement. I was 29 that year. Certainly, I was not thinking of retirement then, but what aroused my interest a few years earlier was whether properties could be a viable investment for me. That led me to study how the demographics of Singapore would change in the 35 years from the time I thought of the question to the time I would retire. My conclusion was the demographics was favourable for the 1st half of the 35 years and unfavourable for the 2nd half of the 35 years. The main issues were the low birth rate and greying population in Singapore. Since I had to live in one house for the entire 35 years (and more), I probably would not profit from any property investment over the entire period. Thus, I dropped the idea of buying a private property. 

In Dec 2004, ST ran a special report on retirement and one of the methods mentioned was to downgrade their homes. Since I had thought about the issue and I believed not many people had considered demographics changes in their planning then, I decided to drop a letter to ST Forum Page. In that letter, I mentioned that due to demographic changes, property sellers would progressively increase relative to buyers and eventually outnumber them in 2017. The analysis behind the letter is discussed in greater details in Properties.

Every 5 years, I would update my analysis based on the latest demographics figures to check if there are any changes to the above conclusion. The last update was in 2013, after the release of the Population White Paper and Land Use Plan (see Properties, the Population White Paper and the Land Use Plan). While the year sellers would begin to outnumber buyers has been pushed further from 2017 till 2020, the broad conclusion from 13 years ago remains unchanged: that sellers will eventually outnumber buyers and property investors are facing a significant headwind from demographics in the long run.

Just a personal note, I believed too much in my own analysis and did not recognise that even if there are no profits to be made over the entire 35-year investment period, it is possible to profit from the 1st half of the period by buying properties or property shares. In fact, I even sold Ho Bee at $0.215 in 2003 and did not pick up Chip Eng Seng in 2003/2004 when it was languishing at $0.10! That is my desserts for being a smart alec! 

The one very expensive lesson that I learnt from this episode is that even though the long-term analysis might point in a certain direction, the short- or medium-term analysis might point in an entirely different direction. I should analyse the short-, medium- and long-term situations instead of being fixated by the long-term conclusion.

See related blog posts:

Sunday, 15 January 2017

Another Year That Ends with 7

You probably have heard of the folklore -- whenever the year ends with 7, the stock market would crash. In 1987, Dow Jones Industrial Average went through the Black Monday in which it crashed 22.6% in a single day on 19 Oct 1987. Stock markets worldwide followed suit. In 1997, Asia went through the Asian Financial Crisis which did not end until nearly 2 years later. The STI went from 2,216.79 on 31 Dec 1996 to 805.04 on 4 Sep 1998, which is a precipitious drop of 64%! Fig. 1 below shows the extent of the crash during that period.

Fig. 1: STI Crash in 1997

In 2007, it was US' turn to experience a financial crisis, which eventually inflicted all other countries in the Global Financial Crisis (GFC). The stock market crash did not begin at the start of 2007, but sometime in Oct 2007 and ended only 1.5 years later. From 3,875.77 on 11 Oct 2007, the STI crashed until it bottomed out at 1,456.95 on 9 Mar 2009 for a steep drop of 62%! Fig. 2 below shows the crash during that period. In fact, 3,875.77 remains the all-time high of STI. For the next 10 years afterwards, the STI never came close to reaching this level.

Fig. 2: STI Crash in 2007

It is 2017 this year, another year that ends with 7. Will history repeat itself and the stock market experience another spectacular crash again?

Personally, I have experienced the crash of 1987, 1997 and 2007. In 1987, I was helping my father to monitor stock prices when he was at work. On that fateful day (20 Oct 1987), I saw prices gapping down by 33% when the market resumed trading from its lunch break. It was such a shock that it became the moment when I knew that the stock market was destined to be a part of my life (see Confessions of a Serious Investor). 

In 1997, I was 1 year away from graduating from university. Being the "smart" guy in the family, I had recommended my father to buy a certain financial stock that had fallen from $3 to $0.75. We never saw the money on this stock again. This crash had the heaviest impact on my family. Because of the financial crisis, Malaysia imposed capital controls, resulting in suspension of trading of all Malaysian stocks listed on the SGX Central Limit Order Book (CLOB) market. All shares were frozen and transferred to the Malaysian stock market. They were only released a few years later. By then, they were mostly worthless. 

In 2007, I was investing with my own money for 9 years when the GFC happened. At the start of 2007, I was uneasy with the speculative fever over structured warrants that pushed stock prices to high levels. Unadjusted for corporate actions, Capitaland reached $8.60, Ezra $6.75, NOL $5.45, SGX $15.40, SIA $19.30, Swiber $3.66! Unfortunately, the money that I pulled out from stocks went into REITs and high-yield business/ shipping trusts, such as FirstShip, Rickmers, MacCookPropSec, etc. which crashed equally significantly. At the depth of the crisis, I estimated I was sitting on paper loss of about 65%! Undaunted, I liquidated half of my bank preference shares and pumped fresh money into the stock market. The market recovered and I recouped all my losses and made some money (see Behind Every Successful Bear Market Recovery is A Cash-Like Instrument).

It is 2017, do I believe in the folklore that the market would crash spectacularly again? I respect folklore, especially having gone through 3 severe market crashes previously, but I needed evidence that a crash is likely, such as sky-high stock prices like the case in 2007. Thus, in the later half of 2016, I was wondering what would cause a crash to materialise. The realisation came when the market did not crash after Brexit happened in Jun 2016: the massive liquidity injected by central banks around the world was propping up asset prices, but that did not help the many companies in many industries which are instead facing poor business and/or low margins, with some companies entering judicial management (see What Have We Got After 8 Years of Easy Money?). Although there is euphoria after the US presidential election that Trump would increase infrastructure spending, cut taxes and regulations and thus speed up the recovery of the economy, he is at the same time advocating protectionist trade policies, potentially triggering trade wars with other countries. Furthermore, increased infrastructure spending would lead to inflation and interest rates rising more rapidly. There is also the risk of capital flight out of non-US countries into US, making US dollar debt burdens more heavy for companies in Asia (see Making America Great Again and Its Impact to Asia). 

Thus, I am not optimistic about the stock market for 2017 and have been shoring up cash positions whenever possible. The crash may or may not happen. But if it happens, I am prepared.

See related blog posts:

Sunday, 8 January 2017

Betting Against the STI Component Stock Changes

Buying and holding a stock index is considered as passive investing. However, the STI is not entirely passive. Every quarter, the STI would be reviewed to determine whether certain stocks should be dropped from or added to the index. This can be likened to active stock selection. How have the stock selections of STI of the past 3 years performed?

Over the past 3 years, there have been 6 changes to the STI component stocks. 1 of them is a forced change, as CapitaMalls Asia (CMA) was privatised and its place in the STI was taken over by Ascendas REIT. All the other 5 changes are voluntary changes. The performance of the stocks removed from the index since their removal till 30 Dec 2016 is shown below.

Stock Replacement Date Price
Price 30/12/16 %
CMA 04 Jun 14 2.35 NA NA A-Reit
JMH 21 Sep 15 47.00 55.25 18% SATS/UOL/YZJ
JSH 21 Sep 15 27.01 33.20 23% SATS/UOL/YZJ
Olam 21 Sep 15 2.04 1.97 -3% SATS/UOL/YZJ
Noble 21 Mar 16 0.22 0.17 -21% CCT
SembMar 19 Sep 16 1.25 1.38 10% JMH


Please note that the price of Noble has been adjusted for a 1-for-1 rights issue at $0.11 in Jun 2016. The unadjusted price before the removal date is $0.32.

Among the 5 stocks removed from the index, 3 have gone on to register double digit gains in stock price, one is esentially flat and another is down by double digits. Overall, this portfolio of discarded index stocks have gained by 5.3% as at 30 Dec 2016.

The performance of the stocks added to the index since their addition till 30 Dec 2016 is shown below.

Stock Replacement Date Price
Price 30/12/16 %
A-Reit 04 Jun 14 2.38 2.27 -5% CMA
SATS 21 Sep 15 3.80 4.85 28% JMH/JSH/Olam
UOL 21 Sep 15 6.02 5.99 0% JMH/JSH/Olam
YZJ 21 Sep 15 1.17 0.81 -30% JMH/JSH/Olam
CCT 21 Mar 16 1.47 1.48 1% Noble
JMH 19 Sep 16 60.14 55.25 -8% SembMar


Among the 6 stocks added to the index, 3 are essentially flat, registering less than 5% gain or loss in stock price. 1 has outperformed with a 28% gain, but it is offset by another which underperformed with a 30% loss. Overall, this portfolio of added index stocks has lost 2.6%.

Please note that this analysis does not consider dividends. Among the 6 added stocks, 2 of them are REITs, which pay handsome dividends but fluctuate little in stock price. This partly explains why the portfolio of added index stocks is essentially flat.

If you had bought the discarded index stocks and sold the added index stocks, you would have gained 5.3% + 2.6% or 7.9%. Note that this is a market-neutral portfolio; you are neither long nor short the market. In other words, this 7.9% gain is alpha! It is not a gain because of a correct bet on the direction of the market. It is a gain independent of where the market moved! (Again, please note that dividends are not considered in this analysis.)

If you notice carefully, there is 1 stock that appears in both the discarded and added index stock lists. The stock is Jardine Matheson Holdings (JMH). When it was dropped from the index in Sep 2015, it price was $47.00. When it was added back to the index a year later, it had risen to $60.14. Its price as at end Dec 2016 was $55.25. Thus, on the 1 occasion when the STI carried out "market-timing", it performed poorly as well!

The main reason for the overperformance of discarded index stocks and underperformance of added index stocks is because changes to the STI component stocks are announced in advance. Hence, investors would have sold the discarded stocks and bought the added stocks before the changes take place, resulting in the discarded stocks outperforming and the added stocks underperforming subsequently. Nevertheless, the fact remains that this is a market inefficiency and investors who bet against the changes stand to profit from it.

In conclusion, the STI is not a very good stock picker. Investors who buy and hold the STI should be aware that changes to the STI component stocks would impact their performance to some extent. Also, betting against the STI component stock changes might be a fairly profitable move.

See related blog posts: