Some musings on portfolio construction based in the UK
Posted: July 31st, 2021, 6:14 pm
Evening All.
Just a few observations, which may be of interest to some.
It seems to me that most forum members tend to report on (and in many case, perhaps construct/manage too) their portfolios in absolute terms, i.e. percentage increases over one or more timeframes - the most common two being since inception or the previous year. There is a variation on this of people who are concerned with their income (as opposed to capital or total return measurements), but for practical purposes, it is still an absolute measurement - just a different kind of one. I suppose this is all somewhat understandable, maybe for three main reasons
I get drawn into these conversations as well from time to time, though I'm not sure how helpful that is, either for me or the person I'm replying to
There are a subset of forum members who are different to this: 1nvest is an obvious one, but there are others as well. The exact form this difference of lens takes varies - for some it is capital preservation, for some it is optimising drawdown within safety parameters (of being wiped out) etc. At some level, I am in this subset - which has led me over the past 12-15 months, to morph my three groups of family portfolios (ISA*2, SIPP*2, JISA*2) from 100% Global Investment Trusts (WTAN, ATST & FRCL respectively) into different, arguably more "risk balanced" holdings. Notwithstanding having worked in the City back in the day, my experience there was more narrow - not really about portfolio management or similar. So, I had working knowledge or better about how most instruments worked, but not necessarily how to use them well together. You might argue I still don't!
So, I had to learn about passive vs active, which products were now on the market to suit each (which led me in part into VWRL and VAGP for the passive component), unitisation etc. But it still wasn't enough, or more accurately, I still couldn't be sure whether what I was doing was at least reasonable. This led me further on the journey - what I really wanted to get was a risk-adjusted feel of my portfolios - without paying too much to get it. To describe my thinking over time in detail would take too long, but five main things I tried were ...
1: Buying "Portfolio Construction", by Russ Koesterlich, on a cheap deal from Harriman House
2: Using PortfolioVisualiser.com to backtest
3: Using PortfolioCharts.com to backtest
4: Using the free version, then taking out a month's free trial with StockCharts.com
5: Getting a MorningStar free account, and trying out its free services
Item (1) proved a decent read, and I would advise most people to consider it or something similar. Were I closer to drawdown, I might instead have purchased McClung's "Living of Your Money" and I probably will do in the future, but I'll wait until I'm closer to retirement and hope for an updated edition with updated data.
Items (2) is good for asset class backtesting, but (understandably) is a bit US centric. That makes its specific instrument backtesting less useful for UK residents.
Item (3) is more useful for UK centric backtesting, as it lets you select your location, but it has less granularity with respect to asset classes, e.g. how to try and get the effect of holding HDIV or BIPS
Item (4) was certainly useful tactically (e.g. it's RRG graphs appeared to help me time some switches that I already wanted to do well) but it didn't have all the UK instruments I needed, or at least their historic detail wasn't available. It has a voting system for new instruments to cover, but this is neither quick nor certain.
Item (5) was of little use, but maybe the paid for version would be, with more on this below.
I then almost decided it was all too hard for a retail UK punter, but have recently come across two things new to me
A: II (Interactive Investor) has some functionality I came across by accident, free with an account, which is almost hidden in the X-Ray (underpinned by MorningStar) capability. There is a little PDF icon in the top right corner of the X-Ray section, strangely unlabelled, which when pressed gives you a much more detailed report - including, like magic, a Correlation Matrix! Suddenly, I can now make use of some of the theory I learned in Portfolio Construction. It will be even more useful when VAGP starts being factored in June 2022 - it requires three years of information for an instrument and it only began in June 2019. One limitation is that you appear not to be able to put in "What If" scenarios - but if anyone else knows how, please let me know?
B: Trustnet's free functionality has a Portfolio section, where, once you've entered in the details, has an "Analysis" tab, which further in has an "Analyse Risk Exposure" tab along with some other useful ones. It considers risk, using FE Fundinfo data (I'm not sure how this compares to Morningstar data) against holding the FTSE100, with cash being 0 and a measurement of 100 equalling the FTSE 100 (for what its worth, my three portfolios measure 64, 73 and 76 respectively). One key thing is that it is UK centric, at least in the version we see - hurrah! As second key thing is that you can specify the period for measurements (which means I can get VAGP data considered if desired).
Of course, this is not the end of the journey, it's likely there isn't one - but with (A) and (B) above and, as I said, even better from June 2022, I now have some tools that are genuinely useful for me. I hope they may be of use to some others.
Regards, Newroad
Just a few observations, which may be of interest to some.
It seems to me that most forum members tend to report on (and in many case, perhaps construct/manage too) their portfolios in absolute terms, i.e. percentage increases over one or more timeframes - the most common two being since inception or the previous year. There is a variation on this of people who are concerned with their income (as opposed to capital or total return measurements), but for practical purposes, it is still an absolute measurement - just a different kind of one. I suppose this is all somewhat understandable, maybe for three main reasons
- It satisfies that natural competitive instincts of many
It's fairly easy to do
How to do something else, e.g. considering it in risk adjusted terms, is not so clear, especially for non-US residents
I get drawn into these conversations as well from time to time, though I'm not sure how helpful that is, either for me or the person I'm replying to
There are a subset of forum members who are different to this: 1nvest is an obvious one, but there are others as well. The exact form this difference of lens takes varies - for some it is capital preservation, for some it is optimising drawdown within safety parameters (of being wiped out) etc. At some level, I am in this subset - which has led me over the past 12-15 months, to morph my three groups of family portfolios (ISA*2, SIPP*2, JISA*2) from 100% Global Investment Trusts (WTAN, ATST & FRCL respectively) into different, arguably more "risk balanced" holdings. Notwithstanding having worked in the City back in the day, my experience there was more narrow - not really about portfolio management or similar. So, I had working knowledge or better about how most instruments worked, but not necessarily how to use them well together. You might argue I still don't!
So, I had to learn about passive vs active, which products were now on the market to suit each (which led me in part into VWRL and VAGP for the passive component), unitisation etc. But it still wasn't enough, or more accurately, I still couldn't be sure whether what I was doing was at least reasonable. This led me further on the journey - what I really wanted to get was a risk-adjusted feel of my portfolios - without paying too much to get it. To describe my thinking over time in detail would take too long, but five main things I tried were ...
1: Buying "Portfolio Construction", by Russ Koesterlich, on a cheap deal from Harriman House
2: Using PortfolioVisualiser.com to backtest
3: Using PortfolioCharts.com to backtest
4: Using the free version, then taking out a month's free trial with StockCharts.com
5: Getting a MorningStar free account, and trying out its free services
Item (1) proved a decent read, and I would advise most people to consider it or something similar. Were I closer to drawdown, I might instead have purchased McClung's "Living of Your Money" and I probably will do in the future, but I'll wait until I'm closer to retirement and hope for an updated edition with updated data.
Items (2) is good for asset class backtesting, but (understandably) is a bit US centric. That makes its specific instrument backtesting less useful for UK residents.
Item (3) is more useful for UK centric backtesting, as it lets you select your location, but it has less granularity with respect to asset classes, e.g. how to try and get the effect of holding HDIV or BIPS
Item (4) was certainly useful tactically (e.g. it's RRG graphs appeared to help me time some switches that I already wanted to do well) but it didn't have all the UK instruments I needed, or at least their historic detail wasn't available. It has a voting system for new instruments to cover, but this is neither quick nor certain.
Item (5) was of little use, but maybe the paid for version would be, with more on this below.
I then almost decided it was all too hard for a retail UK punter, but have recently come across two things new to me
A: II (Interactive Investor) has some functionality I came across by accident, free with an account, which is almost hidden in the X-Ray (underpinned by MorningStar) capability. There is a little PDF icon in the top right corner of the X-Ray section, strangely unlabelled, which when pressed gives you a much more detailed report - including, like magic, a Correlation Matrix! Suddenly, I can now make use of some of the theory I learned in Portfolio Construction. It will be even more useful when VAGP starts being factored in June 2022 - it requires three years of information for an instrument and it only began in June 2019. One limitation is that you appear not to be able to put in "What If" scenarios - but if anyone else knows how, please let me know?
B: Trustnet's free functionality has a Portfolio section, where, once you've entered in the details, has an "Analysis" tab, which further in has an "Analyse Risk Exposure" tab along with some other useful ones. It considers risk, using FE Fundinfo data (I'm not sure how this compares to Morningstar data) against holding the FTSE100, with cash being 0 and a measurement of 100 equalling the FTSE 100 (for what its worth, my three portfolios measure 64, 73 and 76 respectively). One key thing is that it is UK centric, at least in the version we see - hurrah! As second key thing is that you can specify the period for measurements (which means I can get VAGP data considered if desired).
Of course, this is not the end of the journey, it's likely there isn't one - but with (A) and (B) above and, as I said, even better from June 2022, I now have some tools that are genuinely useful for me. I hope they may be of use to some others.
Regards, Newroad