Tatton Investment Management

[Deleted User]

Cus said:

Deleted_User said:

Like John is saying, on a like for like comparison, over 5+ years and assuming we are properly accounting for risk, a portfolio with a 1% extra charge year in, year out is almost certain to underperform. The evidence is overwhelming. And 1% per year is a heck of a lot in this day and age. 

How do you properly account for risk? For example, I pay a wealth management firm to invest for me, charging 1%. It ranges between 60 and 70% equity.  I personally benchmark it against a combination of lifestrategy 60 and 80 ( what I would probably do if I managed it myself)

Currently at 4+ years it's beating those benchmarks significantly. How do I know if my risk level is the same?

You can use measures of volatility, such as Beta or Standard Deviation. They are usually available for a given fund and can be calculated for a portfolio. 

You can also look at the underlying assets.  A typical trick is to compare two 70/30 funds; one of them with government and high quality bonds and the other with much riskier bonds and illiquid assets and claim the former is a benchmark.  

Then you can X ray the equity portion.  Is the average cap size of stocks in your portfolio smaller? That adds risk.  Are you properly diversified internationally? If not -> more risk.  Are you covering all industry sectors or focusing on a couple? More risk.  

cfw1994

Or you could just decide that since it is clearly beating those numbers “significantly”, then you have made good decisions 🤣

[Deleted User]

Its not a 4-year game. 

Sanxxx

Thanks for all the replies, I like the idea of benchmarking against something like Vanguard. Question is, what is the best approach to benchmarking? Is it as simple as 'performance over x years net of charges' ?

I come from an IT background where benchmarking is an art rather than a science and the process is horribly complex with little end value.

cfw1994

Quite a few of us here have an IT background…..& the world of finance, investments & pensions is nothing like a SPECsfs benchmark 🤣

Yes, 'performance over x years net of charges' is the basic measure.  

Of course, you can then bring in “for your risk profile” - Cautious Colin may only hope to match inflation, whilst Boisterous Billy may hope for 5-10% growth YoY.  I perhaps lean more towards the latter, whilst accepting markets can dash those expectations from time to time 🤔


My view on it is to not get hung up on always being The Best: that is a hiding to nothing.  

I focus on the 1/3/5/10 numbers if I am comparing things…but that is more to try to ensure I am not The Worst 🤪
Watch some Lars Kroijer videos for more on that approach.

My other view is keeping things simple….monitor our “net worth” on a quarterly basis - includes property, pensions, ISAs, PBs, etc.  I will use that to frame withdrawals, & the coat will be cut to match the ensuing cloth.

IFAs (& FAs) will have the ability to delve much deeper into this stuff: in part, I view that as all part of maintaining the mystique of financial advice - does it really help, or does it give an illusion of powerful arcane knowledge?  Who knows 😉 

[Deleted User]

Sanxxx said:

Thanks for all the replies, I like the idea of benchmarking against something like Vanguard. Question is, what is the best approach to benchmarking? Is it as simple as 'performance over x years net of charges' ?

I come from an IT background where benchmarking is an art rather than a science and the process is horribly complex with little end value.

The important part is to compare correct “growth” figures.  For the purposes of benchmarking you should use time weighted return.  

Time period is also crucial.  Strategically picking dates can give one the result he/she wants, so should be avoided.  Short periods of time are meaningless. It starts to become meaningful from 10 years and more.  However relative behaviour of various portfoliod during major market crashes  of 2000, 2008 and 2020 is informative. 

Linton

Sanxxx said:

Thanks for all the replies, I like the idea of benchmarking against something like Vanguard. Question is, what is the best approach to benchmarking? Is it as simple as 'performance over x years net of charges' ?

I come from an IT background where benchmarking is an art rather than a science and the process is horribly complex with little end value.

Much the same with investing.  What is different though in investing benchmarks are often used to convince others of the truth of what you already believed to be the case.  There is enough variation in the data and flexibility in its interpretation to prove whatever you want to prove.

What is your objective in benchmarking?  If it is to prove the portfolio A is "better" than portfolio B you are likely to fail to achieve anything useful.
- Unless you make a good effort to ensure that A and B are in some sense comparable you will probably prove what you already knew - eg equities perform better than bonds/cash except when they dont.  What is "comparable"?  Identical is pointless.  There must be some differences in the underlying investments and those differences could be unexpectedly significant.
 - If "better" is purely start to end performance you have the problem that this is highly depedent on events. A different time period or an alternative universe may give different results.  Since you cant predict future events a portfolio unduly dependent on one set of circumstances which happened to occur in your test period cannot sensibly be said to be "better" than one which is more flexible but slightly underperformed.  
 - as Mordko has said, for a benchmark to have some meaning it should be over 10+ years.  Obviously it should include a range of conditions so the absence of a serious crash in the past 10 years makes that period unsuitable.  Did the funds under consideration exist 10 years ago? Were their characteristics the same over that period?  Can you get the data?
 - It would be best to carry out the benchmarking on future data rather than history as it removes any bias arising from your knowledge of past events.  However by the time you have accumulated suficient data to come to a meaningful conclusion it is rather late to do anything about it.

Benchmarking is however useful as it can be very educational.  If one of the portfolios performed much better than the other find out why. Deeper understanding of how investments behave should improve your fund choice in the future.