Mr.Market Simulation

[billyjoe]

I ran a rough simulation of a Mr.Market like portfolio with a few tweaks. Over the years Mr.Market has consistently maintained a winning percentage of over 90%. I tried to err on the conservative side.

1. Maintain positions in 12 stocks at all times.
2. Assume a success rate of 83% (10 of 12 stocks gain 15% before being sold, loss of 20% also triggers sale)
3. Start with equal $$ amounts in each position. In this case, $5000 per stock.
4. Try to keep equal dollar amounts in each position
5. Work with a simple turnover rate of once per year

End of first year, 10 positions $5750 each, 2 positions $4000 each Total $65500
End year 6, 12 positions roughly $8044 each. Total $96532
End year 10, roughly 8 positions $9788, 4 positions $14072 Total $134588
End year 15, 6 positions $13694, 6 positions $19687 Total $200294, a gain of 334%

I think this is a reasonable possibility. What do you think?

--------------billy

[BlueWolf]

Funny, Billy, but I recently did something similar for &%$’s and giggles. In your model, the Compounded Annual Growth Rate (CAGR) ends up at about 8.37. In mine, which were based on a single pass with no double checking, I got slightly better results, with a CAGR of around 10.79, which would make the end number closer to $240,000. In any event, yes, I think your numbers are very believable. There are fund managers that would kill for a CAGR of 10. Of course, in fairness to them, they are managing huge positions, which are much more difficult to trade. Kudos to Mr. Market and thanks for what you do, which is spectacularly generous.

BTW, if you want the details of my methodology and how I got to my number, let me know, and I will post them.

[mrmarket]

Thanks for sharing. Over the last 25 years, I have beaten the S&P 500 in 21 of those 25 years.

[Louetta]

Thanks for sharing. Over the last 25 years, I have beaten the S&P 500 in 21 of those 25 years.

It's interesting that Billy's model includes selling after a 20% loss. Isn't it, Ernie.

[mrmarket]

It's interesting that Billy's model includes selling after a 20% loss. Isn't it, Ernie.

Well...as I have testified all along, and Karel can back me up, if you stay diversified in your portfolio you can stomach a few unlucky plays so you might as well hang on to them if their fundys are good. Having said that, I think for the more risk adverse folks stop losses are a good idea.

[BlueWolf]

It's interesting that Billy's model includes selling after a 20% loss. Isn't it, Ernie.

At one point, I went through the trades and looked at both 15% and 20% stops and found that the use of such stops would have taken you out of a number of winning trades. There might be a stop amount that protects all/most the winning trades, but I don’t know what it is. You’d have to plot the draw down for every trade to compute it. For the losing trades I found, and there weren’t many, the losses were as follows: 64%, 37%, 36%, 92%, and 22%. These percentages were based on estimated closing prices as I could not find the exact closing prices for losing trades. This also doesn’t take into account any gains from dividends. The average loss on losing trades was 50.37%. In any event, you’d have to use a stop of at least 22%, and this could still take you out of winning trades. For example, a 22% stop would have taken you out of the SEDG Trade, which went on to recover and then hit 15%. With a 95% win rate, you can afford to use a very liberal stop, if any at all.

[mrmarket]

At one point, I went through the trades and looked at both 15% and 20% stops and found that the use of such stops would have taken you out of a number of winning trades. There might be a stop amount that protects all/most the winning trades, but I don’t know what it is. You’d have to plot the draw down for every trade to compute it. For the losing trades I found, and there weren’t many, the losses were as follows: 64%, 37%, 36%, 92%, and 22%. These percentages were based on estimated closing prices as I could not find the exact closing prices for losing trades. This also doesn’t take into account any gains from dividends. The average loss on losing trades was 50.37%. In any event, you’d have to use a stop of at least 22%, and this could still take you out of winning trades. For example, a 22% stop would have taken you out of the SEDG Trade, which went on to recover and then hit 15%. With a 95% win rate, you can afford to use a very liberal stop, if any at all.

People have told me to create an ETF but they said that this forum, even though its been around since 2003 would not be an auditable platform for past performance.

[BlueWolf]

OK, nobody asked, but for what it’s worth, here’s the methodology I used. I did not double check my numbers, so there could be a few errors in my raw input. There is also some conservative rounding that would tend to lower the Compounded Annual Growth Rate (CAGR), so keep that in mind. The steps/data I used/computed are as follows:

1) The average hold time, including open trades, was 287 (calendar) days. This number was skewed by losing trades, which were held for much longer periods.

2) The win/loss percentage was about 19/1 or 95%.

3) From July, 2011 to Aug, 2019, a period of 98 months, there were 97 closed trades. For the sake of simplicity, I called that 12 round trip (closed) trades per year. At a 95% win rate, that’s approximately 11 winning trades per year and 1 losing trade. It’s actually 11.4 winning trades, so the rounding here unfairly dilutes the gains of the model.

4) The average gain from winners was 16.35%.

5) The average loss from losers was 50.37%.

6) Additional assumptions and constraints: The maximum number of open positions was 14. However, for the purposes of modeling, since there were an average of 12 trades per year, I used a starting portfolio capable of handling 12 positions of $10,000, i.e., $120,000. Hence there were 12 trades made each year, with the per position allocation equal to 1/12th of the overall portfolio value. This is a simplistic and naive model, but it works for al least getting some idea about CAGR. In the end the CAGR computed out at around 10.79%, which translates to:

Results:
Starting value of portfolio = $120,000.
Year 1, with a per position allocation of $10,000.
End of year value of portfolio = $132,948.
Year 2, with a per position allocation of $11,079.
End of year value of portfolio = $147,292.
Year 3, with a per position allocation of $12,274.
End of year value of portfolio = $163,181.
Year 4, with a per position allocation of $13,598.
End of year value of portfolio = $180,783.
Year 5, with a per position allocation of $15,065.
End of year value of portfolio = $200,286.
etc...
After year 15, the value of the portfolio = approx $558,071.

[mrmarket]

OK, nobody asked, but for what it’s worth, here’s the methodology I used. I did not double check my numbers, so there could be a few errors in my raw input. There is also some conservative rounding that would tend to lower the Compounded Annual Growth Rate (CAGR), so keep that in mind. The steps/data I used/computed are as follows:

1) The average hold time, including open trades, was 287 (calendar) days. This number was skewed by losing trades, which were held for much longer periods.

2) The win/loss percentage was about 19/1 or 95%.

3) From July, 2011 to Aug, 2019, a period of 98 months, there were 97 closed trades. For the sake of simplicity, I called that 12 round trip (closed) trades per year. At a 95% win rate, that’s approximately 11 winning trades per year and 1 losing trade. It’s actually 11.4 winning trades, so the rounding here unfairly dilutes the gains of the model.

4) The average gain from winners was 16.35%.

5) The average loss from losers was 50.37%.

6) Additional assumptions and constraints: The maximum number of open positions was 14. However, for the purposes of modeling, since there were an average of 12 trades per year, I used a starting portfolio capable of handling 12 positions of $10,000, i.e., $120,000. Hence there were 12 trades made each year, with the per position allocation equal to 1/12th of the overall portfolio value. This is a simplistic and naive model, but it works for al least getting some idea about CAGR. In the end the CAGR computed out at around 10.79%, which translates to:

Results:
Starting value of portfolio = $120,000.
Year 1, with a per position allocation of $10,000.
End of year value of portfolio = $132,948.
Year 2, with a per position allocation of $11,079.
End of year value of portfolio = $147,292.
Year 3, with a per position allocation of $12,274.
End of year value of portfolio = $163,181.
Year 4, with a per position allocation of $13,598.
End of year value of portfolio = $180,783.
Year 5, with a per position allocation of $15,065.
End of year value of portfolio = $200,286.
etc...
After year 15, the value of the portfolio = approx $558,071.

Smart people on this forum...for sure.

[jiesen]

OK, nobody asked, but for what it’s worth, here’s the methodology I used. I did not double check my numbers, so there could be a few errors in my raw input. There is also some conservative rounding that would tend to lower the Compounded Annual Growth Rate (CAGR), so keep that in mind. The steps/data I used/computed are as follows:

1) The average hold time, including open trades, was 287 (calendar) days. This number was skewed by losing trades, which were held for much longer periods.

2) The win/loss percentage was about 19/1 or 95%.

3) From July, 2011 to Aug, 2019, a period of 98 months, there were 97 closed trades. For the sake of simplicity, I called that 12 round trip (closed) trades per year. At a 95% win rate, that’s approximately 11 winning trades per year and 1 losing trade. It’s actually 11.4 winning trades, so the rounding here unfairly dilutes the gains of the model.

4) The average gain from winners was 16.35%.

5) The average loss from losers was 50.37%.

6) Additional assumptions and constraints: The maximum number of open positions was 14. However, for the purposes of modeling, since there were an average of 12 trades per year, I used a starting portfolio capable of handling 12 positions of $10,000, i.e., $120,000. Hence there were 12 trades made each year, with the per position allocation equal to 1/12th of the overall portfolio value. This is a simplistic and naive model, but it works for al least getting some idea about CAGR. In the end the CAGR computed out at around 10.79%, which translates to:

Results:
Starting value of portfolio = $120,000.
Year 1, with a per position allocation of $10,000.
End of year value of portfolio = $132,948.
Year 2, with a per position allocation of $11,079.
End of year value of portfolio = $147,292.
Year 3, with a per position allocation of $12,274.
End of year value of portfolio = $163,181.
Year 4, with a per position allocation of $13,598.
End of year value of portfolio = $180,783.
Year 5, with a per position allocation of $15,065.
End of year value of portfolio = $200,286.
etc...
After year 15, the value of the portfolio = approx $558,071.

This was approximately what I calculated some 15 years ago when I first started following $$MM... decided it would be very wise to purchase these picks, and have done so ever since!
(I believe my CAGR on these picks overall has been between 12-14%, which is probably more than most here because I haven't (willingly) sold any of my losers for >50% losses, and many have come back to the + side, even after $$MM has sold these. I have also gone in and out of several picks for a 14% win, rinse, repeat, while $$MM holds out for the 16%+ stretch target on occasion.)
Never been happier with any investment decision I've made more than I am with the one I made when I started buying the $$MM picks!

I simply cannot believe there aren't 1,000,000+ people following this site.