Official Shiny Things thread—Part III

[zwogbwog]

Swan02,

Thank you once again for being so generous with your time and sharing your knowledge with us. I have learnt plenty in the past 48 hours

I can't critique the Reit/physical property plan since I don't have domain expertise.
However, anecdotally, I do use that broad principle with my own portfolio. I would say that about 40-50% of my portfolio is in physical properties/REITS.

From a cash flow perspective, it's really fantastic but I do think that it will really work well if the portfolio is sizeable (>5M).

I was having the same thoughts last night about shifting some of my bond allocation to REITS before I read your post. Not a bad time to buy, given that they have taken a big Covid hammering.

 

[cfleee]

Can someone post the spreadsheet download link from ST on comparing commissions from various brokers? Thanks!

ShinyThings hasn't published any spreadsheet as far as I know, the rules of thumb are all in The Book and supplemented by a ton of replies here... but I do have some spreadsheets modeling that approach of DCAing a monthly income into a 3-fund portfolio, with certain LSE-listed ETFs for the international equity component.

NOTE: These are neither meant as a generic all-purpose broker comparison, nor meant for buying U.S.-listed ETFs. Lots of other comparison pages and tools out there that will deal with those scenarios.

Version 1

v1 was discussed quite thoroughly so it should be reasonably complete. But it doesn't factor in the recent IBKR SG.

Version 2

v2 includes IBKR SG and a bunch of combos that make it look super complicated. I haven't checked this over as thoroughly as I'd like, so I'm not really considering it done yet. Whichever version you use, do your own due diligence.

Also, one known issue is that in v2 I conservatively assumed that all IBKR SG fees and monthly minimum fees would incur 7% GST. It seems that they may not be charging GST (yet?), so this may slightly overestimate the costs for using IBKR SG.

 

[plutofic]

Is it recommended to leverage on investment?

I have extracted the data for IWDA and A35 and calculated their monthly returns from 1 Jun 2012 to 1 Aug 2020. (I use A35 for this study because MBH only started since 2018 and there is not enough data points).

I then calculated the average monthly portfolio returns and standard deviation for these period for the following portfolios:
0% IWDA 100% A35
1% IWDA 99% A35
2% IWDA 98% A35
...
100% IWDA 0% A35
I then plot the graph (in blue) in the link below.
(Note that the standard deviation is a measure of risks)

What I observe is that if I desire a portfolio containing 75% IWDA and 25% A35, I could potentially get a higher return at the same risk if I hold a portfolio of 47% IWDA and 53% A35 at 1.5x leverage. (I assume IBKR margin rate of 1.59% per annum which is around 0.13% per month)

Do you all think this is a good idea? Or I am missing something here?

Apologies, I do not have sufficient post count to post an image. Hence, would need you to copy and paste the link below on your browser to see the image
imgur.com/a/pxmwtrF

For your information, here are the data points:
0% IWDA 100% A35: Returns 0.22% STD 0.0139
10% IWDA 90% A35: Returns 0.30% STD 0.0131
20% IWDA 80% A35: Returns 0.38% STD 0.0139
30% IWDA 70% A35: Returns 0.46% STD 0.0161
40% IWDA 60% A35: Returns 0.54% STD 0.0191
47% IWDA 53% A35: Returns 0.60% STD 0.0216
50% IWDA 50% A35: Returns 0.62% STD 0.0227
60% IWDA 40% A35: Returns 0.70% STD 0.0266
70% IWDA 30% A35: Returns 0.78% STD 0.0307
75% IWDA 25% A35: Returns 0.82% STD 0.0328
80% IWDA 20% A35: Returns 0.86% STD 0.0349
90% IWDA 10% A35: Returns 0.94% STD 0.0393
100% IWDA 0% A35: Returns 1.02% STD 0.0436

 

[martypants]

To be frank, there is no safe way to get a 6.6% yield out of investment-grade SGD bonds, no matter how hard you leverage them. You’re taking a lot of risk; my advice would be to lower your expectations. (Hey, I’m just being honest, because your RM isn’t.)

(Also, LQDA and SDIA own USD-denominated bonds, so they have currency risk as well. They’re not the appropriate investment for someone in your situation.)

If you need $1500 a month, safely, you can get that by buying $800k worth of MBH. That’s safe, and sensible, and it won’t go wrong and you won’t have to pay tens of thousands a year in hidden fees to do it.

Seems like I need to go back to the drawing board here. Thanks for such a generous answer Shiny!

 

[hwckhs]

I applaud you for taking the effort to do numerical research and getting your hands dirty doing calculations. You will always learn something from such exercises.

I am just trying to understand your ideas. So am asking questions rather than giving advice.

I have extracted the data for IWDA and A35 and calculated their monthly returns from 1 Jun 2012 to 1 Aug 2020. (I use A35 for this study because MBH only started since 2018 and there is not enough data points).

I assume you use the ETFs' prices. Did you consider dividends (of A35), and did you convert IWDA to SGD to that they have a common base currency?

What I observe is that if I desire a portfolio containing 75% IWDA and 25% A35, I could potentially get a higher return at the same risk if I hold a portfolio of 47% IWDA and 53% A35 at 1.5x leverage. (I assume IBKR margin rate of 1.59% per annum which is around 0.13% per month)

How do you get a higher return with the same risk? Can you show the calculation?

47% IWDA 53% A35: Returns 0.60% STD 0.0216
75% IWDA 25% A35: Returns 0.82% STD 0.0328

You used the period "1 Jun 2012 to 1 Aug 2020". Will the same observation holds if you select a different period, ie. selection bias? My gut feeling is that it probably will not. Correct me if I'm wrong.

Furthermore, the margin rate is likely variable and not constant throughout the period?

 

[plutofic]

I applaud you for taking the effort to do numerical research and getting your hands dirty doing calculations. You will always learn something from such exercises.

Yes indeed, I started out trying to find the optimal portfolio allocation and noticed this.

I assume you use the ETFs' prices. Did you consider dividends (of A35), and did you convert IWDA to SGD to that they have a common base currency?

Yes, I have added the dividend for A35 and convert IWDA to SGD using the exchange rate at the respective days.

How do you get a higher return with the same risk? Can you show the calculation?

I only draw lines on the graphs and did not calculate exactly. Shall do it now:
Monthly margin rate = 1.0159^(1/12)=1.00132
Gradient of graph = (0.0060-0.00132)/(0.0216-0)=0.217
(y-0.00132)=0.217(x-0) --> y=0.217x+0.00132
At STD of 0.0328, returns=0.217*0.0328+0.00132=0.0084
Hence, leverage with 47% IWDA 53% A35 gives 0.84% compared to 0.82% with a 75% IWDA 25% A35 portfolio.

You used the period "1 Jun 2012 to 1 Aug 2020". Will the same observation holds if you select a different period, ie. selection bias? My gut feeling is that it probably will not. Correct me if I'm wrong.

I selected 1 Jun 2012 because it is the furthest I could go. Yahoo finance gives weird data prior to Jun 2012.
I just experimented with different period of data. Here are my observation:
- For most period, the graph is similar shape so this would work
- For recent months, A35 returns is higher than IWDA and the graph become "inverted". This method performed way better than 75/25 portfolio!
- For few periods where A35 return was way below margin rate, the 75/25 portfolio would performed better.

Furthermore, the margin rate is likely variable and not constant throughout the period?

This is something I have not considered. If margin rate increase beyond 2.4% a year, the 75/25 portfolio would performed better in this model compared to leveraged.

 

[hwckhs]

I just experimented with different period of data. Here are my observation:
- For most period, the graph is similar shape so this would work
- For recent months, A35 returns is higher than IWDA and the graph become "inverted". This method performed way better than 75/25 portfolio!
- For few periods where A35 return was way below margin rate, the 75/25 portfolio would performed better.

...

This is something I have not considered. If margin rate increase beyond 2.4% a year, the 75/25 portfolio would performed better in this model compared to leveraged.

It sounds like there is no guarantee that leverage will generate a better return than non-leverage. The only guarantee I'm seeing is that the broker will earn from the margin rate .

 

[hwckhs]

@plutofic, leverage is for short term play. Never heard of ppl doing that for long term investment.

Googled a bit and found https://www.reddit.com/r/investing/comments/blz8y5/why_would_a_long_term_investor_not_use_margin/ which contains all the arguments against leveraging long-term. If you are doing long-term, may just forget leveraging and focus on AA instead.

 

[tcwehcs]

Hi ST what’s your view on high yield bonds ETF like HYG and JNK? And if you know if there is 30% withholding tax apply for their month distributions?

 

[plutofic]

It sounds like there is no guarantee that leverage will generate a better return than non-leverage. The only guarantee I'm seeing is that the broker will earn from the margin rate .

Yes, there is no guarantee. Just trying to find ways to increase the probability of a higher returns.

@plutofic, leverage is for short term play. Never heard of ppl doing that for long term investment.

Googled a bit and found (URL removed because I could not post URL) which contains all the arguments against leveraging long-term. If you are doing long-term, may just forget leveraging and focus on AA instead.

I am not very familiar with margin investing. People in that reddit thread were talking about 3% margin rate. I guess the 1.59% margin rate that I see now is due to low interest rate environment and may not be sustainable in the long term. If the rate increase to 3%, I will be worse off leveraging.

Regarding margin call:
I just read up a bit on it. For 25% maintenance margin requirement, and 1.5x leverage, the price would have to drop 55.6% before triggering a margin call. It is unlikely for A35 to drop this much but its a possibility for IWDA during deep financial crisis. It would be detrimental when margin call occurs as we will be forced to sell at low price.

The risk, which is represented by standard deviation did not capture the risk of margin call and changing margin rate. You are right that I should focus on AA instead. Will stick to 50% IWDA, 25% ES3, 25% MBH, though I am considering increasing IWDA and decreasing ES3.

 

[zoneguard]

Yes, there is no guarantee. Just trying to find ways to increase the probability of a higher returns.

The authors of this book talked about using LEAPs instead of margin trading as a form of leveraged diversification over time.
There's some data at the site if you wish to look into the details.

For those interested in all-weather portfolio, SGX published an article on their take with STI/A35/GLD and the YTD/3 years returns.

 

[swan02]

I'm not used to see STD in monthly.

1. How do you annualize your STD ? do you multiply your STD monthly by the square root of 12 ?

2.also I find STD expressed as relative STD easier to understand. Is this expressed as so ?

3. I just want to ensure I understand your linear graph as I'm quite confused. Noticed you have made some changes to an error.

4. I would have expected that the optimal AA would have to be substantially lower something like 30/70 before leverage is applied. However considering that a35 is not as safe haven as usd treasury, your AA might be correct.

5. You look like a student of Quantitative finance. Whereas I'm a student of Dr google finance.

6. I'm at best math mediocre. I'm trying to learn without taking a CFA course.

thanks.

I only draw lines on the graphs and did not calculate exactly. Shall do it now:
Monthly margin rate = 1.0159^(1/12)=1.00132
Gradient of graph = (0.0060-0.00132)/(0.0216-0)=0.217
(y-0.00132)=0.217(x-0) --> y=0.217x+0.00132
At STD of 0.0328, returns=0.217*0.0328+0.00132=0.0084
Hence, leverage with 47% IWDA 53% A35 gives 0.84% compared to 0.82% with a 75% IWDA 25% A35 portfolio.

 

[plutofic]

The authors of this book talked about using LEAPs instead of margin trading as a form of leveraged diversification over time.
There's some data at the site if you wish to look into the details.

Looks interesting! I have not read the full book but from the excerpt, seem like they were talking about using call options. Instead of margin call, options have expiry dates. If the option expires when it is "out of the money" (when market price is lower than strike price), the person will lose 100% of the money used to buy the option. But I think the book probably recommend a combination of options and long/short stocks/bonds to mitigate this risk but it will reduce returns or/and substitute existing risks to other risks. I will explore further when I can find the time. Have you tried their strategy? How does it fare?

 

[plutofic]

I'm not used to see STD in monthly.

1. How do you annualize your STD ? do you multiply your STD monthly by the square root of 12 ?

2.also I find STD expressed as relative STD easier to understand. Is this expressed as so ?

3. I just want to ensure I understand your linear graph as I'm quite confused. Noticed you have made some changes to an error.

4. I would have expected that the optimal AA would have to be substantially lower something like 30/70 before leverage is applied. However considering that a35 is not as safe haven as usd treasury, your AA might be correct.

5. You look like a student of Quantitative finance. Whereas I'm a student of Dr google finance.

6. I'm at best math mediocre. I'm trying to learn without taking a CFA course.

thanks.

1. Statistically, I think so. But I think you would get a different number if you calculate annual STD directly.

2. No, this is just standard deviation. I think the Markowitz Portfolio Theory uses standard deviation instead of relative STD.

3. Don't worry, it was just a typo when I copied here. The graph is still the same.

4. The graph will shift if we use T-bills (lower returns and lower STD). But this seem to suggest lower bond weightage on the optimal AA.
Also, if the margin rate is lower, then the optimal AA bond weightage will be greater!

5. I was an engineering major! I just happened to take a few finance modules when I was in university, out of curiosity! I am working in a non finance related job so I am just doing investment on my own, following mainly what was discussed in this thread and hoping to be rich by retirement

 

[zoneguard]

Have you tried their strategy? How does it fare?

There are no IWDA/VWRA options/LEAPs to implement the strategy with. I don't want to use US options/LEAPs due to tax considerations. I appreciate the authors' concept of temporal diversification and its benefits but execution is challenging in our context.

 

[plutofic]

There are no IWDA/VWRA options/LEAPs to implement the strategy with. I don't want to use US options/LEAPs due to tax considerations. I appreciate the authors' concept of temporal diversification and its benefits but execution is challenging in our context.

Yeah many of the investment books are written for USA context, especially some which contain elements of tax avoidance strategies. But in some sense, we are fortunate that our taxation is more straight forward.

 

[swan02]

I simply annualised your monthly STD using the said formula which then made a lot more sense to me. Had a look at your graph.

1. I was actually referring to 7-10 year Treasury bonds not bills.

2. From my understanding, seeing that the tangent portfolio requires such little A35, tells me that A35 isn't as great a diversifier as say 7-10 yr Treasury Bonds. This isn't surprising as it confirms my suspicions when holding to a not as good safe haven govt bond (A35). Its tardiness is likely also due to the SGD exposure.

It also means, I need to take a greater unit of risk in IWDA (risk on) to achieve an optimal level of return should I continue to stick purely on A35 hence I wonder how the tangent portfolio would look like once I mix with USD treasury bonds.


To put in another way, it says I will be penalised with suboptimal returns once I start having A35 beyond 53 percent in order to match a reduced risk appetite. How ironic.


3. On the positives, at least now we know we need 53 percent of A35 and 47 percent IWDA to enjoy the best sharpe ratio, and lever at this portfolio as long we can find means to a low cost of funds guaranteed for long periods, and a way of not being exposed to margin calls.

Perhaps I might be motivated to NOT pay off my HDB loan. I wonder 2.6 percent cost of funds (HDB) guaranteed is worth levering onto this portfolio ?
I think you would be motivated to find out ?

4. well I have to ensure your data sets are accurate, time for me to practice and learn.

5. I wonder r u interested to see how MBH fares even with a short data period ?, at least it went through the recent crisis.

6. Lastly, can we have the correlation the next time you do your calculations ?

1. Statistically, I think so. But I think you would get a different number if you calculate annual STD directly.

2. No, this is just standard deviation. I think the Markowitz Portfolio Theory uses standard deviation instead of relative STD.

3. Don't worry, it was just a typo when I copied here. The graph is still the same.

4. The graph will shift if we use T-bills (lower returns and lower STD). But this seem to suggest lower bond weightage on the optimal AA.
Also, if the margin rate is lower, then the optimal AA bond weightage will be greater!

5. I was an engineering major! I just happened to take a few finance modules when I was in university, out of curiosity! I am working in a non finance related job so I am just doing investment on my own, following mainly what was discussed in this thread and hoping to be rich by retirement

 

[perrinrahl]

@Shiny and BBC

The recovery of IWDA and to a lesser extent VWRD from pandemic lows has been remarkable. Cf. local indices still in the doldrums and looking to be that way for some time to come.

While this may a short-term trend due to the pandemic, what are your thoughts wrt the weightage between local and global ETFs? I know BBC favours a much larger ratio for the latter. Even before Covid19, I have started to weight more towards IWDA, if not exactly at BBC's ratio.

What are your thoughts?

 

[Ms.FindingMoney]

Hello people,

I am looking for a BioTech ETF. I found ARKG but I am wondering if there are other ETFs with a lower expense ratio? Additionally, I am not sure about the tax on this ETF. I understand we buy IWDA to reduce tax from 30 to 15%.

For context, I am already invested in IWDA / ES3. Just looking for other areas to invest in.

 

[kevinkoh1992]

In view of unlimited QE, should asset allocation be modified to ride out stagflation?