How To Forecast Portfolio Loss?

[steadyeddy]

Assume a bear market with a 20% & 30% broad stock market correction.

For each, have you calculated what your portfolio would lose?

Please share your thoughts on how to calculate that?

Thanks.

Eddy

[Mustang]

Since I believe that any method used will be somewhat inaccurate I would use something simple. I'd look at history. I often use the difference in the 2008 losses between Wellington and Wellesley Income to explain why I want less volatility and am currently buying Wellesley. That type of comparison could be expanded. For example:
....... Vanguard 500 Index (VFINX) -37.03%  (times .9 for portion of portfolio = 33.57%)

....... Vanguard Mid-Cap Index (VIMSX) -41.82%  (times .1 for portion of portfolio = 4.18%)

....... Weighted Avg loss = 37.75%

My three largest funds' 2008 performance was:
....... American Funds Balanced Fund (ABALX) -25.73%.  This is 65% of portfolio.
....... Vanguard Wellington (VWENX) -22.23%.  This is 24% of portfolio.
....... Vanguard Wellesley Income (VWINX) -9.84%.  This is 11% of portfolio.

The mid-cap portion of stocks in my funds is very small. Wellington has almost no mid-cap. Wellesley maybe 10%. And Balanced maybe 13%. That probably explains a little of why Balanced's loss was greater than Wellington's. Well, that and the fact that balanced is more growth oriented. Even though the difference is small I thought I should use a weighted average 2008 market loss. That is shown above.

So for a 30% broad market loss I would guess that:
... Balanced would lose 30 x (25.73/37.75) = 20.45 x 65% of portfolio = 13.29%
... Wellington would lose 30 x (22.23/37.75) = 17.67 x 24% of portfolio = 4.24%
... Wellesley would lose 30 x (9.84/37.75) = 7.81 x 11% of portfolio = 0.86%

Estimate portfolio loss for a 30% drop in market is the total of 18.39%.

But this is where withdrawal strategies come into play. I will be using RMDs (a dynamic withdrawal strategy)from Balanced Fund so my dollar withdrawal will be 20.45% less than it was the year before leaving more shares behind for the recovery. And using my methodology the withdrawal from my taxable funds will come from Wellesley not Wellington. Leaving shares in Wellington for the recovery.

I'm sure computers can come up with a more precise number out to ten decimal places which would also be wrong. Off the top of my head I would have guessed the loss would be two-thirds the market drop. That would have been a little high but probably close enough for any guess about the future.

P.S. That was kind of fun.  A lot better than watching re-runs.  I hope I didn't screw it up too badly.

[yogibearbull]

I use effective-equity [= Relative SD wrt SP500].

With so much StimulusN and easy Fed, I don't see a bear market in 2021, but may be a +/- 10% boring year. I am aware that historically, the 1 yr in Presidential cycle tends to be bad as Presidents try to clean house to get ready for the 2nd yr that is mid-term Congressional election yr, and then for the 4th yr that is for their own preservation/incumbency.

My effective-equity is around 50% in retirement [target range 40-60%], so I am looking for boring +/- 5% in 2021. But if I am wrong, and 20-30% bear came, my portfolio will temporarily decline [not lose] 10-15%.

If the current bull run continues, I may be reducing my effective-equity to 40-45% range, and then even if that 20-30% bear comes, my decline would be even less.

[steadyeddy]

Mar 8, 2021 1:03:20 GMT Mustang said:

Since I believe that any method used will be somewhat inaccurate I would use something simple. I'd look at history. I often use the difference in the 2008 losses between Wellington and Wellesley Income to explain why I want less volatility and am currently buying Wellesley. That type of comparison could be expanded. For example:
....... Vanguard 500 Index (VFINX) -37.03%  (times .9 for portion of portfolio = 33.57%)

....... Vanguard Mid-Cap Index (VIMSX) -41.82%  (times .1 for portion of portfolio = 4.18%)

....... Weighted Avg loss = 37.75%

My three largest funds' 2008 performance was:
....... American Funds Balanced Fund (ABALX) -25.73%.  This is 65% of portfolio.
....... Vanguard Wellington (VWENX) -22.23%.  This is 24% of portfolio.
....... Vanguard Wellesley Income (VWINX) -9.84%.  This is 11% of portfolio.

The mid-cap portion of stocks in my funds is very small. Wellington has almost no mid-cap. Wellesley maybe 10%. And Balanced maybe 13%. That probably explains a little of why Balanced's loss was greater than Wellington's. Well, that and the fact that balanced is more growth oriented. Even though the difference is small I thought I should use a weighted average 2008 market loss. That is shown above.

So for a 30% broad market loss I would guess that:
... Balanced would lose 30 x (25.73/37.75) = 20.45 x 65% of portfolio = 13.29%
... Wellington would lose 30 x (22.23/37.75) = 17.67 x 24% of portfolio = 4.24%
... Wellesley would lose 30 x (9.84/37.75) = 7.81 x 11% of portfolio = 0.86%

Estimate portfolio loss for a 30% drop in market is the total of 18.39%.

But this is where withdrawal strategies come into play. I will be using RMDs (a dynamic withdrawal strategy)from Balanced Fund so my dollar withdrawal will be 20.45% less than it was the year before leaving more shares behind for the recovery. And using my methodology the withdrawal from my taxable funds will come from Wellesley not Wellington. Leaving shares in Wellington for the recovery.

I'm sure computers can come up with a more precise number out to ten decimal places which would also be wrong. Off the top of my head I would have guessed the loss would be two-thirds the market drop. That would have been a little high but probably close enough for any guess about the future.

P.S. That was kind of fun.  A lot better than watching re-runs.  I hope I didn't screw it up too badly.

Mustang - Thank you for taking the time to illustrate how you calculated potential (paper) loss if a bear struck. Wellesley is my largest holding too.

[steadyeddy]

Mar 8, 2021 2:05:29 GMT yogibearbull said:

I use effective-equity [= Relative SD wrt SP500].

With so much StimulusN and easy Fed, I don't see a bear market in 2021, but may be a +/- 10% boring year. I am aware that historically, the 1 yr in Presidential cycle tends to be bad as Presidents try to clean house to get ready for the 2nd yr that is mid-term Congressional election yr, and then for the 4th yr that is for their own preservation/incumbency.

My effective-equity is around 50% in retirement [target range 40-60%], so I am looking for boring +/- 5% in 2021. But if I am wrong, and 20-30% bear came, my portfolio will temporarily decline [not lose] 10-15%.

If the current bull run continues, I may be reducing my effective-equity to 40-45% range, and then even if that 20-30% bear comes, my decline would be even less.

YBB - You explained effective equity once before. Is there a tool or excel formula to compute this for an entire portfolio?

[yogibearbull]

Mar 8, 2021 3:14:31 GMT steadyeddy said:

Mar 8, 2021 2:05:29 GMT yogibearbull said:

I use effective-equity [= Relative SD wrt SP500].

With so much StimulusN and easy Fed, I don't see a bear market in 2021, but may be a +/- 10% boring year. I am aware that historically, the 1 yr in Presidential cycle tends to be bad as Presidents try to clean house to get ready for the 2nd yr that is mid-term Congressional election yr, and then for the 4th yr that is for their own preservation/incumbency.

My effective-equity is around 50% in retirement [target range 40-60%], so I am looking for boring +/- 5% in 2021. But if I am wrong, and 20-30% bear came, my portfolio will temporarily decline [not lose] 10-15%.

If the current bull run continues, I may be reducing my effective-equity to 40-45% range, and then even if that 20-30% bear comes, my decline would be even less.

YBB - You explained effective equity once before. Is there a tool or excel formula to compute this for an entire portfolio?

PV run provides data for it. See details in LINK.

[steadyeddy]

Mar 8, 2021 3:21:15 GMT yogibearbull said:

Mar 8, 2021 3:14:31 GMT steadyeddy said:

YBB - You explained effective equity once before. Is there a tool or excel formula to compute this for an entire portfolio?

PV run provides data for it. See details in LINK.

Thank YOU, yogibearbull.

[Mustang]

yogibearbull , I went to portfolio visualizer and put in my funds using VWELX instead of VWENX because it has a longer history. It was very easy to use especially when I only had three funds. It said the standard deviation of my portfolio was 9.11%. How does that tell me what the expected loss will be if the general broad market loses 30%?

Out of curiosity I also used it having only one fund, Vanguard Wellesley Income (VWINX). It has a standard deviation of 6.61%. How does that help me forecast the market decline of Wellesley Income if the broad market loses 30%? I already know that in 2008 a SP500 index fund lost 37% (the SP 500 dropped 38.5%) and Wellesley lost 9.8%. How does portfolio visualizer tell me that will happen?

[yogibearbull]

Mar 8, 2021 13:23:24 GMT Mustang said:

yogibearbull , I went to portfolio visualizer and put in my funds using VWELX instead of VWENX because it has a longer history. It was very easy to use especially when I only had three funds. It said the standard deviation of my portfolio was 9.11%. How does that tell me what the expected loss will be if the general broad market loses 30%?

Out of curiosity I also used it having only one fund, Vanguard Wellesley Income (VWINX). It has a standard deviation of 6.61%. How does that help me forecast the market decline of Wellesley Income if the broad market loses 30%? I already know that in 2008 a SP500 index fund lost 37% (the SP 500 dropped 38.5%) and Wellesley lost 9.8%. How does portfolio visualizer tell me that will happen?

To the PV run, add SP500 as benchmark, or VFINX or SPY as an additional portfolio, to get SD_sp500 for the same timeframe as for the others. Then find the ratio SD_portfolio/SD_sp500. Don't use SDs from different periods or sources. Use month-to-month settings, so recent 3 years would mean March 2018 to February 2021; PV uses 1st day of the start month and last day of the end month.

[ignatz]

Mar 8, 2021 3:14:31 GMT steadyeddy said:

Mar 8, 2021 2:05:29 GMT yogibearbull said:

I use effective-equity [= Relative SD wrt SP500].

With so much StimulusN and easy Fed, I don't see a bear market in 2021, but may be a +/- 10% boring year. I am aware that historically, the 1 yr in Presidential cycle tends to be bad as Presidents try to clean house to get ready for the 2nd yr that is mid-term Congressional election yr, and then for the 4th yr that is for their own preservation/incumbency.

My effective-equity is around 50% in retirement [target range 40-60%], so I am looking for boring +/- 5% in 2021. But if I am wrong, and 20-30% bear came, my portfolio will temporarily decline [not lose] 10-15%.

If the current bull run continues, I may be reducing my effective-equity to 40-45% range, and then even if that 20-30% bear comes, my decline would be even less.

YBB - You explained effective equity once before. Is there a tool or excel formula to compute this for an entire portfolio?

Yes.

You can use STDEV in Excel to calculate the standard deviation over your choice of time periods. You can use weekly or monthly figures....whatever you have records for.

I use monthly and a 3 year period....for my portfolio and for the SP500. It's simple to then calculate how one compares the other.

Right now, the 3 year SD of my portfolio is about 75 percent of the SD of the SP500. My portfolio is about 61 percent equities.

The action in Feb 2020 had a significant effect on the SD.

My calculated comparison matches that shown by Portfolio Visualizer.

HOWEVER....history has shown me that expected declines are typically underestimated...at least once a decade. And that once a decade is usually for a big decline.

So, you spend a jillion hours estimating that you will lose 18 percent and in fact lose 24 in real life. Or, you have a 40/60 portfolio and lose 60 percent of the SP's loss. Stuff like that. A battleship gray to charcoal swan rears its head.

[yogibearbull]

Mar 8, 2021 13:58:16 GMT ignatz said:

Mar 8, 2021 3:14:31 GMT steadyeddy said:

YBB - You explained effective equity once before. Is there a tool or excel formula to compute this for an entire portfolio?

Yes.

You can use STDEV in Excel to calculate the standard deviation over your choice of time periods. You can use weekly or monthly figures....whatever you have records for.

I use monthly and a 3 year period....for my portfolio and for the SP500. It's simple to then calculate how one compares the other.

Right now, the 3 year SD of my portfolio is about 75 percent of the SD of the SP500. My portfolio is about 61 percent equities.

The action in Feb 2020 had a significant effect on the SD.

My calculated comparison matches that shown by Portfolio Visualizer.

HOWEVER....history has shown me that expected declines are typically underestimated...at least once a decade. And that once a decade is usually for a big decline.

So, you spend a jillion hours estimating that you will lose 18 percent and in fact lose 24 in real life. Or, you have a 40/60 portfolio and lose 60 percent of the SP's loss. Stuff like that. A battleship gray to charcoal swan rears its head.

Great that you are keeping monthly portfolio records.

You are probably aware that SDs reported by PV, M*, etc are for monthly total returns, not for values [NAVs or balances]. For others, PV remains a great free tool [still]. 

If your nominal-equity is 61%, but Relative SD [effective-equity] is 75%, it is the latter that matters.

I have tested Relative SDs over various timeframes and they are quite stable.  If the effective-equity is 18% and actual decline is 24% [or, the reverse], that is within the ballpark. But I haven't encountered a case where effective-equity is 40% but the actual decline is 60%. On the other hand, there are many examples where nominal-equity is 40%, but actual decline is 60%, and effective-equity would provide a better picture. I have posted elsewhere that, based on effective-equity, several conservative-allocation funds are acting like moderate-allocation, and moderate-allocation acting like aggressive-allocation, and that their holders should at least be aware of that, and not be surprised.

[Mustang]

Thanks for the information. The standard deviation for VFINX for the same time frame is 15.12%. 9.11/15.12 = .6025. So for a 30% broad market drop I'd expect an 18.18% drop in my portfolio. My back of the envelope calculation was 18.39%. Pretty close.

But here is why both are simply WAGs (wild ... guesses). I also estimated that Wellesley Income would lose 7.8%. The portfolio visualizer method estimates a(6.61/15.12) = 13.12% loss. They are very different and here is a little bit of why. I used only 2008. Portfolio visualizer uses all of the years from 1985. Here are some of the SP 500 losses during that time frame. Please note Wellesley Income earned profits during three of them. Considering current bond performance I think the more recent data would result in the most accurate estimate.

Year........ SP500.......VWINX
2018....... -6.24%...... -2.57%
2008...... -38.37%...... -9.84%
2002...... -23.37%...... +4.64%
2001...... -13.44%...... +7.39%

2000...... -10.14%..... +16.17%

For a diversified portfolio the SP 500 is the wrong benchmark.  I would think the greater the proportion of bonds the further it would be off.  It's like comparing apples to oranges.

[ignatz]

You can certainly tune the SD formula in Excel to include whatever time frame you want, as long as you have returns for that period.

I don't often look at Portfolio Visualizer, but I would have thought you can control the time frame there as well.

Likewise, you could use VT or VTI or whatever fund or index you want to use for comparison purposes....rather than the SP.

My primary point would be that you'd be on very shaky ground to "expect" any particular decline...regardless of what it is based on or how it is calculated. So-called aberrations in returns occur more frequently and have more impact than typically anticipated. Being out near the end of the SD curve is a far cry from impossible and those things pop up all too often....and that can be quite damaging in declines. I think we have "100 year floods" every decade nowadays.

[steadyeddy]

Mar 8, 2021 17:08:13 GMT Mustang said:

Thanks for the information. The standard deviation for VFINX for the same time frame is 15.12%. 9.11/15.12 = .6025. So for a 30% broad market drop I'd expect an 18.18% drop in my portfolio. My back of the envelope calculation was 18.39%. Pretty close.

But here is why both are simply WAGs (wild ... guesses). I also estimated that Wellesley Income would lose 7.8%. The portfolio visualizer method estimates a(6.61/15.12) = 13.12% loss. They are very different and here is a little bit of why. I used only 2008. Portfolio visualizer uses all of the years from 1985. Here are some of the SP 500 losses during that time frame. Please not Wellesley Income earned profits during three of them. Considering current bond performance I think the more recent data would result in the most accurate estimate.

Year........ SP500.......VWINX
2018....... -6.24%...... -2.57%
2008...... -38.37%...... -9.84%
2002...... -23.37%...... +4.64%
2001...... -13.44%...... +7.39%

2000...... -10.14%..... +16.17%

For a diversified portfolio the SP 500 is the wrong benchmark.  I would think the greater the proportion of bonds the further it would be off.  It's like comparing apples to oranges.

Mustang, thanks for posting this info. Wellesley is of immense interest to me since it is the largest holding. The higher the market goes the more I add to Wellesley...

[Chahta]

Taking into account how bond funds behave, how does Effective Equity work for a total portfolio? Bonds can go down with stocks, go up against stocks, dive like they did in the last black swan or drop in a correction like bonds did the last few weeks. Still the SD for many bond funds is high due to the black swan. When we speak of market decline we are speaking about the equity market.

If using PV Backtest, then MAX Drawdown will indicate what your decline was for FEB/MAR 2020 and verify your Effective Equity.

[Mustang]

Mar 9, 2021 13:07:42 GMT Chahta said:

Taking into account how bond funds behave, how does Effective Equity work for a total portfolio? Bonds can go down with stocks, go up against stocks, dive like they did in the last black swan. Still the SD for many bond funds is high due to the black swan.

It could be argued that any of the large market crashes of the past were black swans.  If the predictive tools had seen them coming investors would have changed their behavior avoiding the crash.  Investopedia labels both 2008 and 2001 as black swans. www.investopedia.com/terms/b/blackswan.asp

I remember "Black Monday" (1987) where computer programs caused the crash.  That was definitely a black swan. corporatefinanceinstitute.com/resources/knowledge/trading-investing/black-monday/

So if you can't predict it what can you do?  It depends upon where you are in your investment life cycle.  In 1987 I owned Vanguard Wellington.  I was 37 years old and far from retirement.  I pretty much emptied my savings account buying more.  In 2008 I was 58 and retirement was within sight.  I was probably 80-85% in stock funds.  That was when I decided to tone back my stock exposure.  When I retired from my company in 2010 and transferred my 401k to a traditional IRA I did exactly that with most going into American Funds Balanced Fund.  I didn't even notice 2018.  We were not yet withdrawing from our portfolio.  But I was starting to look at the various withdrawal methods.  Working through that convinced me I needed more stability in my taxable investments so we started buying Wellesley Income.

If history is any indicator the market recovers from crashes.  The key to surviving it is selling as little as possible while the market is down.  That is hard to do when you are living off retirement investments.  But there are alternatives other than bonds.  Some advisors recommend two or even three years of withdrawals in cash.  Other things like annuities and reverse mortgages can provide cash flow allowing you to avoid selling of shares during the crash.  Wellesley Income Fund is also being used for that purpose although it is less effective than some of the others.

This is where a withdrawal plan comes into play.  How you choose to withdraw money can dictate the type of investments you need.

[steadyeddy]

Mar 9, 2021 13:49:18 GMT Mustang said:

Mar 9, 2021 13:07:42 GMT Chahta said:

Taking into account how bond funds behave, how does Effective Equity work for a total portfolio? Bonds can go down with stocks, go up against stocks, dive like they did in the last black swan. Still the SD for many bond funds is high due to the black swan.

It could be argued that any of the large market crashes of the past were black swans.  If the predictive tools had seen them coming investors would have changed their behavior avoiding the crash.  Investopedia labels both 2008 and 2001 as black swans. www.investopedia.com/terms/b/blackswan.asp

I remember "Black Monday" (1987) where computer programs caused the crash.  That was definitely a black swan. corporatefinanceinstitute.com/resources/knowledge/trading-investing/black-monday/

So if you can't predict it what can you do?  It depends upon where you are in your investment life cycle.  In 1987 I owned Vanguard Wellington.  I was 37 years old and far from retirement.  I pretty much emptied my savings account buying more.  In 2008 I was 58 and retirement was within sight.  I was probably 80-85% in stock funds.  That was when I decided to tone back my stock exposure.  When I retired from my company in 2010 and transferred my 401k to a traditional IRA I did exactly that with most going into American Funds Balanced Fund.  I didn't even notice 2018.  We were not yet withdrawing from our portfolio.  But I was starting to look at the various withdrawal methods.  Working through that convinced me I needed more stability in my taxable investments so we started buying Wellesley Income.

If history is any indicator the market recovers from crashes.  The key to surviving it is selling as little as possible while the market is down.  That is hard to do when you are living off retirement investments.  But there are alternatives other than bonds.  Some advisors recommend two or even three years of withdrawals in cash.  Other things like annuities and reverse mortgages can provide cash flow allowing you to avoid selling of shares during the crash.  Wellesley Income Fund is also being used for that purpose although it is less effective than some of the others.

This is where a withdrawal plan comes into play.  How you choose to withdraw money can dictate the type of investments you need.

Mustang, words of wisdom from experience. Thank You! I definitely intend to have a cash buffer for a few years regardless of my AA. And the AA is fairly conservative as well.

[FD1000]

Mar 7, 2021 19:22:59 GMT steadyeddy said:

Assume a bear market with a 20% & 30% broad stock market correction.

For each, have you calculated what your portfolio would lose?

Please share your thoughts on how to calculate that?

Thanks.

Eddy

Pretty difficult to predict when you use several categories.  The simpler your portfolio, the easier it gets.  Example: use SPY/VTI for stocks and VBTLX=BND for bonds. Then assume all the losses from stocks and no gain from bonds. 
To check the above you can use March 2020.  VBINX is 60/40 and VFIAX=SP500.
Last year from 2/20/2020 to 3/23/20...VFIAX -33.55%...VBINX -22.7%...VBTLX -1.1%     60% of 33.55 = 20.1 and you add the -1.1 loss and it's not far from 22% loss.
So, to be sure use 65-70% of your stocks % as max loss.

But wait, when you use managed and more flexible funds you can't make accurate prediction but you can still guesstimate.

Myself? I always make assumptions based on worse case scenarios.

But, the only way I know how to avoid big losses is to sell to cash.  Timing is not recommended for most investors but it worked pretty well for me. I lost less than 1% in Q4/2020 and none in March 2020.

Attachments:

[galeno]

Our 45/55 port would decline by less 4% and 9% for equity crashes of 20 and 30% respectively.