Wednesday, September 23, 2020

Barbells, Pyramids And Your Portfolio

If you have been investing for some time, you may have heard or read about some other portfolio strategies. Two that were often mentioned would be the barbell portfolio strategy and pyramid investing strategy, which I will call them barbell and pyramid respectively.

Most investment portfolios and funds come in the form of a pie chart, denoting the various asset classes, geographical locations, sectors and industrials, companies, and their respective make-up, usually in percentages. Some argued that such charts do not necessarily convey an important component of portfolio management, and that is risk. I can show a pie chart that is made up of 60% tech stocks that had gone way above their valuations and 40% junk bonds, but to the novice (or even a seasoned investor if he/she is not careful) the risks are not clearly defined. The barbell and pyramid enable a viewer to see the risk factor of each financial instrument roughly in their respective places.


Let us take a closer look at these two.




In the mainstream barbell portfolio strategy, securities and financial instruments are grouped into either low-to-no-risk or high-risk categories, and there is no middle ground. An example would be to have government bonds and defensive counters at the low-to-no-risk part, while placing those highly valued tech stocks on the other. If a 50-50 split is adopted, it would look something like this:


We could see an asymmetric-looking barbell, depending on the risk appetite and profile of the investor. If an investor prefers more safe instruments than riskier ones, then the bottom triangle would look larger; or for an aggressive, high-risk taker, the top would be larger instead.


Some variations of the barbell included a middle ground, with medium-risk securities making up the middle (e.g. corporate bonds), forming the “handle” of the barbell. This portion kind of give a “best of both worlds” scenario of having counters between the risk spectrum.




The pyramid investing strategy, or sometimes called the risk pyramid, labelled investments into three main tiers: low-risk forming the wide base (typically 40% to 50%), medium-risk as the body (30% to 40%) and high-risk being the apex (10% to 30%), as shown below:


Usually low-risk assets such as cash, government bonds, etc. are at the bottom, followed by the typical dividend-paying equities, corporate bonds, REITs, etc. in the middle, and volatile counters at the top. In some combinations, the pyramid is further split into more than three, especially in the middle part where the medium is further classified into lower-medium, medium-medium or upper-medium, much like our current description of the current social middle class (e.g. lower-upper middle class).


Side Note: Defining Risk


The very first issue on using such strategies would be the definition of risk, since it is this very parameter that you are classifying your investments. Risk is subjective: it means different things to different people at different times. We could say about government bonds being low risk, since it is after all the debt of a country, but different countries have different risk profiles, and some provide yields that are comparable to those of our local REITs’. Also, some counters which are deemed defensive in nature years ago may not appear to be so now (case in point: a telco counter that had went to a 12-year low recently).


We could use numbers to define them, such as standard deviation on measuring volatility, a typically used risk metric, or ratios like Sharpe and Sortino. The main problem in using them is they are not easily found online, and even if available, they may not be free, especially figures of local counters. The other way is to calculate them out, and spare time and Excel know-how is needed.


Another easier way is to use the beta coefficient (or beta for short), which is the measure of a security’s return relative to the general market. With the market’s beta as 1, a counter with a lower beta is seen as less volatile than the market. Beta numbers are easily searched; Yahoo Finance has them. The problem of beta is due to its relativity to the market returns, and a low beta does not mean it is not volatile by itself.


In addition, these statistics are based on the past, not the future, so the risk definition is at best backward looking. While it is understandable to use past data as an indication, it is not foolproof, but we have to make do with whatever information is at hand at least to have a clear set of criteria.


Application On Investment Portfolios


Summarizing my views, which I would elaborate below, in terms of visualizing risk, the barbell is good for looking at one portfolio, while the pyramid is more for looking at multiple portfolios and a risk overview of your total finances.


Let us take a look at the barbell. Using the balanced Bedokian Portfolio (35% equities, 35% REITs, 20% bonds, 5% commodities and 5% cash), I would place the cash and government bonds at the low-to-no risk portion; corporate bonds, low volatile/defensive equities and REITs at the middle (by the way I am using the handle model, which I will state why later); and the higher volatile/growth counters and commodities at the top. If I have other portfolios using my Central Provident Fund, Supplementary Retirement Scheme monies and/or with my disposal income (investment or trading), I can convert them to barbells, too.


With the few barbells, I will create a pyramid by corresponding their risk levels from each barbell: the high-risk parts will form the apex, the financial instruments in the handles will form the middle (here is the reason why I used the handle model), and the low-to-no-risk parts at the base. To make your pyramid more holistic, you may want to include all other holdings outside of your investment and trading portfolios, such as your emergency funds, cash savings, insurance savings plans, etc.




The barbell and pyramid models work well for investors who diversify their portfolios by risk and/or follow the simple equity-bond portfolio. As for myself, I will view them as risk-perspective models that I can transform my pie-chart to for a better understanding. However, I still hold my diversification belief in the order of asset classes, regions/countries, sectors/industries and individual companies and organisations, that would reduce risk.

Sunday, September 20, 2020

Inside The Bedokian’s Portfolio: Keppel DC REIT

Inside The Bedokian’s Portfolio is an intermittent series where I will reveal what we have in our investment portfolio, one company/bond/REIT/ETF at a time. In each post I will briefly give an overview of the counter, why I had selected it and what possibly lies ahead in its future.

For this issue, we will look at Keppel DC REIT.




Keppel DC REIT (KDC) was (and still is) the only locally listed, dedicated data centre REIT. Listed in 2014, it began with eight data centres in Singapore, Malaysia, Australia, United Kingdom, Netherlands and Ireland. Fast forward to 2020, KDC had expanded, holding 19 assets spread across eight countries. 


Let us look at some key statistics of KDC1, 2, based on its closing price of SGD 3.03 on 18 Feb 2020:

  • Trailing Dividend Yield: 2.66%
  • Price-to-Book Value: 2.59
  • Gearing: 34.5%
  • Weighted Average Lease Expiry: 7.4 years


Why KDC?


Back in 2014 while researching on the viability and future trends of E-commerce using associative investing, I had identified a few areas where I could tap onto this potential, and one of them was data centres. The opportunity came around that time when KDC was about to be listed and I had managed to secure some shares at the IPO phase.


Over the past six years, the developmental direction of the internet was not just moving along the E-commerce route, but other areas as well, such as cloud storages, the explosion of various social media platforms, and recently due to COVID-19, the push for the mass adoption of digital and virtual solutions. It was also mainly because of these factors that KDC showed resiliency, and its price went to an all-high time, despite suffering a slouch during the market’s dive back in March 2020.


The Future


With the current trends that I am observing, internet-based technologies and usages (including those that were mentioned in the previous paragraph) are on an upward trajectory, and data centres (and subsequently the REITs that are owning them) will benefit from this. Proving their importance, Mapletree Industrial Trust had completed its acquisition of their remaining interest in 14 data centres in the United States3, and is looking to acquire one more4, setting themselves for the potentiality of data centres.


There are challenges to be faced by data centres themselves, majority of which comes from the things that keep them going. Infrastructure issues like power management and efficient cooling need to be addressed constantly. Couple this with the increasing awareness of being green (data centres are known to be power guzzlers), these could form make-or-break factors in a tenant’s selection criteria.


Nevertheless, data centres in general (not just KDC’s) are the way to go.




Bought Keppel DC REIT at: 


SGD 0.93, Dec 2014 (IPO)

SGD 1.155, Nov 2016 (Rights issue)

SGD 1.71, Sep 2019 (Preferential offer)




1 – Yahoo Finance. Keppel DC REIT. 18 Feb 2020. (accessed 19 Feb 2020)


2 – Keppel DC REIT. First Half 2020 Financial Results. 21 July 2020. (accessed 19 Feb 2020)


3 – Mui, Rachel. Mapletree Industrial Trust completes acquisition of 14 data centres in US. The Business Times. 2 Sep 2020. (accessed 19 Feb 2020)


4 – Mui, Rachel. Mapletree Industrial Trust to buy US data centre for up to US$262.1m. The Business Times. 14 Sep 2020. (accessed 19 Feb 2020)

Monday, September 7, 2020

Frasers Centrepoint Trust’s Acquisition Of AsiaRetail Fund’s Singapore Assets

In what is to be one of the recent biggest news in the REIT scene, Frasers Centrepoint Trust (FCT) is going to acquire the remaining 63.1% of the AsiaRetail Fund (ARF) assets in Singapore, which comprises of Century Square and Tampines 1 in Tampines, Hougang Mall in Hougang, Tiong Bahru Plaza and Central Plaza in Tiong Bahru, and White Sands in Pasir Ris. Also announced in the news is the divestment of Bedok Point to their sponsor Frasers Property for about SGD 108 million.

Finance wise, FCT is proposing to raise about SGD 1.39 billion via equity, in the form of private placement and/or preferential offering, to fund the acquisition and to repay existing debt. Post-acquisition and divestment, on a pro forma basis there would be an 8.59% distribution per unit accretion and the net asset value of FCT would be at SGD 2.21.


The acquisition also marks FCT’s first foray into offices in the form of Central Plaza, which contributed about 4.9% of the net property income (NPI) of the ARF Singapore assets in the financial year (FY) of 2019.


More information can be found in the official presentation slides here.


The Bedokian’s Take


What interested me is the two maps which were shared in the abovementioned presentation slides, which showed the catchment area (or “location of influence” in my write up here) of the malls (Figure 1) and their proximity to the MRT stations (or “location of complements” in the same said write up) (Figure 2). I had just written a piece about FCT a while back here where I had described them as the King of the North. With the full acquisition of the ARF properties, FCT is poised to have a huge retail influence and foothold not just the population centres in the north, but in the north-east and east, too.

Fig.1 - 3km catchment (or radius) of FCT and ARF malls

Fig.2 - Locations of FCT and ARF malls and their walking distances to the nearest MRT stations

Despite competition coming in for retail businesses in the form of online shopping and e-commerce, and the recent COVID-19 situation which saw a decline in footfall due to prevention measures such as circuit breaker and social distancing, the crowds are still there. Just visit any shopping centre (FCT’s or others’) during the peak periods and you can see for yourself. The additional advantage of FCT’s malls is that most of them are situated at the residential heartlands, which traditionally act as a hub for nearby home dwellers to run their errands, have their meals (dine-in or take-away) and/or fulfilling other social and recreational needs such as meet-ups. Malls and shopping centres, in our local context, do fill a special place in our way of life. Factoring in the new phenomenon of work from home, I do see potential in this enlarged REIT.


The divestment of Bedok Point is seen as a good move, since it was not ideally located due to its proximity to the much bigger and nearer-to-the-MRT-station Bedok Mall (owned by CapitaLand Mall Trust), and it contributed only about 1.91% and 1.85% of FCT’s NPI in FY2019 and FY2018 respectively (note: NPI did not include those from Waterway Point and ARF malls)1


An extraordinary general meeting will be convened on 28 September 2020 to pass the resolutions which include the proposed fund raising and the divestment of Bedok Point.




The Bedokian is vested in FCT.


1 – FCT Annual Report 2019 p31. (accessed 6 Sep 2020)


Further references:


FCT Circular dated 3 Sep 2020. (accessed 6 Sep 2020)


Lee, Marissa. Frasers Centrepoint Trust to raise up to S$1.39b to take over AsiaRetail Fund. Business Times. 3 Sep 2020. (accessed 6 Sep 2020)