How to Estimate Potential Losses

[Deleted User]

kempiejon said:

I assume everything on a traded market could drop by 50% or more, perhaps less. Can I find uncorrelated assets and diversify within and without different markets.
I don't think I have the nous or inclination to calculate a risk number for my holdings or what I'd do with it if I looked up ratios someone else has calculated.
 It looks like the DC metric is specific to active funds, an interesting idea so thanks for that link

The down-market capture ratio measures how well an investment manager performs relative to a benchmark index during periods of market decline.

How well a fund tracks an index I would have thought of as tracking error but I've not spent long looking at active funds to see how they fare.

Yes, this is for managed funds only. I only buy managed funds and have sold all but one remaining tracker. The reason quite simply is because at this point in time, trackers offer zero downside protection in a market where downside risk is high and increasing and at my age, I may not have enough investing years remaining, to recoup any losses. The tracker I continue to hold is the FTSE All World, which in itself is a diluted US index that I sold recently and even that is held at no more than 25%.

HedgehogRulez

Sounds like there’s a massive correction coming to global markets soon!

Best buy tins of sprouts now whilst they’re cheap!

[Deleted User]

masonic said:

chiang_mai said:

A few examples of funds below showing MS risk level, Beta, DS ratio and volatility. The Artemis SG GEMS fund is higher risk yet the DC ratio is only 59, meaning, it could only lose half as much as the index in the event of a markets fall whilst index tracker funds such as FTSE AW may lose 80%. 

Those metrics are only relevant under the conditions present over the period they were determined. Beware recency bias, as often this is just a few years. I tend to use: how did it, or a similar fund, perform in the dotcom crash, GFC, and Covid crash as a rough yardstick. Even then, one must be cautious that approaches that limited downside in specific crashes may not work next time.

In the case of Artemis SG GEMS, it fell peak to trough 27% vs 31% for Vanguard GEMs, so I think considering it to have substantially lower loss potential than the index is rather optimistic. It has also very closely tracked the index over the course of 2025, including the tariff correction (-10% in both cases).

Yes, the cause and nature of any market falls is a variable which may negate any benefit of diversification and change the behaviour of funds performance.

In the examples I posted, M&G Japan show a high Beta and high DC ratio yet the the fund is primarily large and giant caps. Polar Japan however is the opposite, it is low Beta, low DC yet it is comprised of a majority of small caps. Interestingly, both funds have very similar MS risk ratings. During market falls, small caps generally lose more value more quickly than their Large cap counterparts so exactly what would happen there during a crash, is very uncertain. I imagine the small cap fund would be hit hardest, which would make the low DC ratio argument meaningless...dunno!

[Deleted User]

HedgehogRulez said:

Sounds like there’s a massive correction coming to global markets soon!

Best buy tins of sprouts now whilst they’re cheap!

I don't think either one of your statements is necessarily correct, the price of tinned sprouts for example is near its peak.

InvesterJones

chiang_mai said:

masonic said:

chiang_mai said:

A few examples of funds below showing MS risk level, Beta, DS ratio and volatility. The Artemis SG GEMS fund is higher risk yet the DC ratio is only 59, meaning, it could only lose half as much as the index in the event of a markets fall whilst index tracker funds such as FTSE AW may lose 80%. 

Those metrics are only relevant under the conditions present over the period they were determined. Beware recency bias, as often this is just a few years. I tend to use: how did it, or a similar fund, perform in the dotcom crash, GFC, and Covid crash as a rough yardstick. Even then, one must be cautious that approaches that limited downside in specific crashes may not work next time.

In the case of Artemis SG GEMS, it fell peak to trough 27% vs 31% for Vanguard GEMs, so I think considering it to have substantially lower loss potential than the index is rather optimistic. It has also very closely tracked the index over the course of 2025, including the tariff correction (-10% in both cases).

Yes, the cause and nature of any market falls is a variable which may negate any benefit of diversification and change the behaviour of funds performance.

In the examples I posted, M&G Japan show a high Beta and high DC ratio yet the the fund is primarily large and giant caps. Polar Japan however is the opposite, it is low Beta, low DC yet it is comprised of a majority of small caps. Interestingly, both funds have very similar MS risk ratings. During market falls, small caps generally lose more value more quickly than their Large cap counterparts so exactly what would happen there during a crash, is very uncertain. I imagine the small cap fund would be hit hardest, which would make the low DC ratio argument meaningless...dunno!

Won't the beta etc. have been calculated from historical data? So the statement 'small caps generally lose more value more quickly than their Large cap counterparts' can't be correct, at least for the data used to calculate the stat. Unless you're talking about a type of fall that hasn't been seen in the data yet.

masonic

InvesterJones said:

chiang_mai said:

masonic said:

chiang_mai said:

A few examples of funds below showing MS risk level, Beta, DS ratio and volatility. The Artemis SG GEMS fund is higher risk yet the DC ratio is only 59, meaning, it could only lose half as much as the index in the event of a markets fall whilst index tracker funds such as FTSE AW may lose 80%. 

Those metrics are only relevant under the conditions present over the period they were determined. Beware recency bias, as often this is just a few years. I tend to use: how did it, or a similar fund, perform in the dotcom crash, GFC, and Covid crash as a rough yardstick. Even then, one must be cautious that approaches that limited downside in specific crashes may not work next time.

In the case of Artemis SG GEMS, it fell peak to trough 27% vs 31% for Vanguard GEMs, so I think considering it to have substantially lower loss potential than the index is rather optimistic. It has also very closely tracked the index over the course of 2025, including the tariff correction (-10% in both cases).

Yes, the cause and nature of any market falls is a variable which may negate any benefit of diversification and change the behaviour of funds performance.

In the examples I posted, M&G Japan show a high Beta and high DC ratio yet the the fund is primarily large and giant caps. Polar Japan however is the opposite, it is low Beta, low DC yet it is comprised of a majority of small caps. Interestingly, both funds have very similar MS risk ratings. During market falls, small caps generally lose more value more quickly than their Large cap counterparts so exactly what would happen there during a crash, is very uncertain. I imagine the small cap fund would be hit hardest, which would make the low DC ratio argument meaningless...dunno!

Won't the beta etc. have been calculated from historical data? So the statement 'small caps generally lose more value more quickly than their Large cap counterparts' can't be correct, at least for the data used to calculate the stat. Unless you're talking about a type of fall that hasn't been seen in the data yet.

Low beta can come about through low liquidity. If a share isn't traded often, as can be the case for certain small caps, then it may not move as much as large cap counterparts on typical days. Since beta is calculated from daily price movements, this can mask a greater loss capacity that fades into the statistical background during linear regression. The fall might have been seen, but it has a low weighting in the data.

[Deleted User]

InvesterJones said:

chiang_mai said:

Yes, the cause and nature of any market falls is a variable which may negate any benefit of diversification and change the behaviour of funds performance.

In the examples I posted, M&G Japan show a high Beta and high DC ratio yet the the fund is primarily large and giant caps. Polar Japan however is the opposite, it is low Beta, low DC yet it is comprised of a majority of small caps. Interestingly, both funds have very similar MS risk ratings. During market falls, small caps generally lose more value more quickly than their Large cap counterparts so exactly what would happen there during a crash, is very uncertain. I imagine the small cap fund would be hit hardest, which would make the low DC ratio argument meaningless...dunno!

Won't the beta etc. have been calculated from historical data? So the statement 'small caps generally lose more value more quickly than their Large cap counterparts' can't be correct, at least for the data used to calculate the stat. Unless you're talking about a type of fall that hasn't been seen in the data yet.

That's the dichotomy, as a broad general rule, small caps are the first to fall and they usually fall the farthest....I can atest to that being true from watching VG Small Caps plummet on cue on many occasions. Yet depsite that apparent "norm", Polar Japan's low DC ratio suggest that hasn't happened with their fund.....which either means the rule doesn't hold true all the time or that there is something different in their case (as masonic suggests in the previous post).  

OldScientist

chiang_mai said:

A few examples of funds below showing MS risk level, Beta, DS ratio and volatility. The Artemis SG GEMS fund is higher risk yet the DC ratio is only 59, meaning, it could only lose half as much as the index in the event of a markets fall whilst index tracker funds such as FTSE AW may lose 80%. 

Ms Risk	Beta	Equities Funds	D/S Cap	Volatility
87	0.61	Art SmartGARP EU Equities	-9	9.7%

Just to add to the comment made by @masonic - the period over which the measurement is made is crucial. To take the above fund as an example, the downside capture ratio was 2, 51, and 101 over 3, 5, and 10 years, respectively. I also note that the maximum drawdown over 10 years for this fund (-35%) was greater than for the index (-21%) that morningstar are comparing it against.

In terms of estimating the effect of sustained crashes - testing the robustness of your plan to a variety of drops is probably useful. For example, perhaps 2-5 year crashes of 40%, 50%, 60% for the equity component and 20% for the bond component (depending on the duration of the funds you have) give a fairly nasty test, but not out of line with worst case historical tests (e.g., US 1929 and Japan 1989 onwards).

IMV, trying to predict how individual active equity funds will respond during a crash is a fool's errand, although active bond funds may be more predictable at least in nominal terms (since duration is known, although noting that corporate bonds tend to be more equity-like than government issues).

[Deleted User]

OldScientist said:

chiang_mai said:

A few examples of funds below showing MS risk level, Beta, DS ratio and volatility. The Artemis SG GEMS fund is higher risk yet the DC ratio is only 59, meaning, it could only lose half as much as the index in the event of a markets fall whilst index tracker funds such as FTSE AW may lose 80%. 

Ms Risk	Beta	Equities Funds	D/S Cap	Volatility
87	0.61	Art SmartGARP EU Equities	-9	9.7%

Just to add to the comment made by @masonic - the period over which the measurement is made is crucial. To take the above fund as an example, the downside capture ratio was 2, 51, and 101 over 3, 5, and 10 years, respectively. I also note that the maximum drawdown over 10 years for this fund (-35%) was greater than for the index (-21%) that morningstar are comparing it against.

In terms of estimating the effect of sustained crashes - testing the robustness of your plan to a variety of drops is probably useful. For example, perhaps 2-5 year crashes of 40%, 50%, 60% for the equity component and 20% for the bond component (depending on the duration of the funds you have) give a fairly nasty test, but not out of line with worst case historical tests (e.g., US 1929 and Japan 1989 onwards).

IMV, trying to predict how individual active equity funds will respond during a crash is a fool's errand, although active bond funds may be more predictable at least in nominal terms (since duration is known, although noting that corporate bonds tend to be more equity-like than government issues).

Thank you and @masonic for exploring this issue although I'm afraid doing so has raised more questions than answers. Yes, the measurement timeframe will be important because the composition of the fund will change over time, as will the economic environements within which the fund exists. And since not all loss contributing factors are necessarily present during every market downturn, attempting to understand the usefullness of DC ratio's is questionable, without significant and frequent research. An attractive DC ratio during one financial year, may well mask factors hitherto unseen in other years, which, when those factors come into play, produce a totally different picture. I am therefore left wondering what the usefullness is of equities funds DC ratio's, other than as a snapshot in time.

But what I did get out of this discussion is the need to evaluate performance over longer periods. I didn't think that that a few years might be regarded as recency but I can now understand the argument that it might be. 

I'm also left struggling to understand what metrics other people use to assess the risk of their investment portfolio. Capitalisation ratio's, geographic spreads, sector utilisation and concentration levels are all good common sence indicators. Volatility readings are also helpful but i don't think proprietary risk level readings are since they are relative to their peers rather than to the market or indicies. Beta is useful to a degree as a reflection of volatility but not that helpful in determining potential loss. Perhaps, at the end of the day, the only meaningful indicator is the hi-level asset allocation and the extent to which a person is invested in equities versus other less risky options.....dunno.    

Back to the drawing board. 

Linton

IMHO You really need to accept that the future is unknown. No amount of statistics is going to help predict what will happen in a major crash initiated by some external event.

My metric is simply the % allocation to any particular category of possibly correlated equities. If you minimise this you will get some protection barring a complete global collapse.