backtesting different indexes

Have a question about your personal investments? No matter how simple or complex, you can ask it here.
indexonlyplease
Posts: 526
Joined: Thu Apr 30, 2015 12:30 pm
Location: Pembroke Pines, FL

backtesting different indexes

Postby indexonlyplease » Mon Jul 17, 2017 5:56 am

Remember just asking question for education:

I am backtesting different indexes on portfolioviser. Leaving bonds out. Only stocks. From 1972-2017 I did 3 possible basic portfolios: each portfolio was 50% large, 25% mid and 25% small. Not sure why I picked the %. Are these unrealistic %.

large, mid, small cap
large, mid, small growth
large, mid, small value

value best, then cap and last was growth. I understand we have no idea of the future. But I do undestand most do backtesting to check possible portfolio performance.

I can undestand no going with all value but does the cap blend seems to be good. I think if I added the total stock it does not go back to 1972.

Can you tell where I am wrong. What missing on my thinking.

I have to add this. When I go back to 1972 what a big difference. But when I go back to 1986 (my investing start) there is no crazy difference. So is this just because from 1986 on the returns were so close. And going foward could be the same??? of course unknown.

bloom2708
Posts: 2240
Joined: Wed Apr 02, 2014 2:08 pm
Location: Fargo, ND

Re: backtesting different indexes

Postby bloom2708 » Mon Jul 17, 2017 2:01 pm

Backtesting has limited value. Don't get hooked on analysis paralysis. Nobody knows if the future will be like the past.

Start with Total US + Total International.

If you want to take a "flier" on small cap value or emerging value or some other tilt, then add an additional fund to execute the tilt.
"We are here not to please but to provoke thoughtfulness" Unknown Boglehead | | Want to buy something? Watch this first: https://vimeo.com/41152287

alex_686
Posts: 2005
Joined: Mon Feb 09, 2015 2:39 pm

Re: backtesting different indexes

Postby alex_686 » Mon Jul 17, 2017 2:09 pm

indexonlyplease wrote:value best, then cap and last was growth. I understand we have no idea of the future. But I do undestand most do backtesting to check possible portfolio performance.


Most people start out with a investment thesis. Then they do back testing, being aware of its limits.

So my question to you is this - why do you think value outperforms? What is the underlying economic reason for this anonymity, why do you think this will persist, is this exploitable, and how confident are you of your analysis?

There are underlying market structures and these structures change. These change. The past may have under valued but I think that anonymity has ended.

indexonlyplease
Posts: 526
Joined: Thu Apr 30, 2015 12:30 pm
Location: Pembroke Pines, FL

Re: backtesting different indexes

Postby indexonlyplease » Mon Jul 17, 2017 3:47 pm

bloom2708 wrote:Backtesting has limited value. Don't get hooked on analysis paralysis. Nobody knows if the future will be like the past.

Start with Total US + Total International.

If you want to take a "flier" on small cap value or emerging value or some other tilt, then add an additional fund to execute the tilt.


I am already sold on the 3 fund portfolio I have. I am just playing with the backtesting and asking questions. Yes it does look like someone should own value funds because of past performance but we know nothing about the future.

indexonlyplease
Posts: 526
Joined: Thu Apr 30, 2015 12:30 pm
Location: Pembroke Pines, FL

Re: backtesting different indexes

Postby indexonlyplease » Mon Jul 17, 2017 3:48 pm

alex_686 wrote:
indexonlyplease wrote:value best, then cap and last was growth. I understand we have no idea of the future. But I do undestand most do backtesting to check possible portfolio performance.


Most people start out with a investment thesis. Then they do back testing, being aware of its limits.

So my question to you is this - why do you think value outperforms? What is the underlying economic reason for this anonymity, why do you think this will persist, is this exploitable, and how confident are you of your analysis?

There are underlying market structures and these structures change. These change. The past may have under valued but I think that anonymity has ended.


I have no basis because I have no idea what will happen in the future. Just playing with backtesting and asking questions.

pkcrafter
Posts: 11670
Joined: Sun Mar 04, 2007 12:19 pm
Location: CA
Contact:

Re: backtesting different indexes

Postby pkcrafter » Mon Jul 17, 2017 4:10 pm

indexonly, I just happened to come across some small cap performance data that may interest you. I posted it in Theory, and here's a direct link to the data.

https://www.raymondjames.com/melvillewe ... rmance.pdf


Paul
When times are good, investors tend to forget about risk and focus on opportunity. When times are bad, investors tend to forget about opportunity and focus on risk.

alex_686
Posts: 2005
Joined: Mon Feb 09, 2015 2:39 pm

Re: backtesting different indexes

Postby alex_686 » Tue Jul 18, 2017 9:07 am

indexonlyplease wrote:I have no basis because I have no idea what will happen in the future. Just playing with backtesting and asking questions.


First, read these 2 links.
https://xkcd.com/882/
https://fivethirtyeight.com/features/sc ... ken/#part1

Back? O.K. Finding interesting things is actually very interesting. Finding significant things is harder.

Second, read up on Fama and French 3 factor model, which is basically what you have got. No reason to reinvent the wheel.

https://www.bogleheads.org/wiki/Fama_an ... ctor_model

I think Fama and French has been supplanted by Carhart's four-factor model, which adds momentum.

Third, there is hefty discussion on what these factors mean. These factors tend to wax and wane over time. Value may be in ascendance one year, growth the next. If I recall correctly if you want to exploit these factors you need a time horizon of 10 years.

Are they real? For example, it looks like the value factor delivers higher returns. However it also might hide higher risks. Value companies declare bankruptcy at a higher rate than growth companies.

Will these factors hold? One suggestion is that value outpreforms growth because they are hidden pearls. Investors tend to focus on sexy companies that get lots of media coverage - that is growth companies. However now that investors have figured out the value of value they have piled into the value space. No more hidden pearls.

If I recall correctly, the value outpreform factor has been dropping. Is this a natural waning or a fundamental shift in market factors?

Does this help?

indexonlyplease
Posts: 526
Joined: Thu Apr 30, 2015 12:30 pm
Location: Pembroke Pines, FL

Re: backtesting different indexes

Postby indexonlyplease » Tue Jul 18, 2017 9:23 am

alex_686 wrote:
indexonlyplease wrote:I have no basis because I have no idea what will happen in the future. Just playing with backtesting and asking questions.


First, read these 2 links.
https://xkcd.com/882/
https://fivethirtyeight.com/features/sc ... ken/#part1

Back? O.K. Finding interesting things is actually very interesting. Finding significant things is harder.

Second, read up on Fama and French 3 factor model, which is basically what you have got. No reason to reinvent the wheel.

https://www.bogleheads.org/wiki/Fama_an ... ctor_model

I think Fama and French has been supplanted by Carhart's four-factor model, which adds momentum.

Third, there is hefty discussion on what these factors mean. These factors tend to wax and wane over time. Value may be in ascendance one year, growth the next. If I recall correctly if you want to exploit these factors you need a time horizon of 10 years.

Are they real? For example, it looks like the value factor delivers higher returns. However it also might hide higher risks. Value companies declare bankruptcy at a higher rate than growth companies.

Will these factors hold? One suggestion is that value outpreforms growth because they are hidden pearls. Investors tend to focus on sexy companies that get lots of media coverage - that is growth companies. However now that investors have figured out the value of value they have piled into the value space. No more hidden pearls.

If I recall correctly, the value outpreform factor has been dropping. Is this a natural waning or a fundamental shift in market factors?

Does this help?


I will review all the articles. But just thinking out loud since someone does not know if growth or value will occur in the future, this would be the reason for blended fund. Instead of large growth large blend etc. Looks this way when backtesting. Or even better the total fund for the sector. If that makes sense.

indexonlyplease
Posts: 526
Joined: Thu Apr 30, 2015 12:30 pm
Location: Pembroke Pines, FL

Re: backtesting different indexes

Postby indexonlyplease » Tue Jul 18, 2017 9:24 am

pkcrafter wrote:indexonly, I just happened to come across some small cap performance data that may interest you. I posted it in Theory, and here's a direct link to the data.

https://www.raymondjames.com/melvillewe ... rmance.pdf


Paul


Willl read.

Thanks

alex_686
Posts: 2005
Joined: Mon Feb 09, 2015 2:39 pm

Re: backtesting different indexes

Postby alex_686 » Tue Jul 18, 2017 10:13 am

indexonlyplease wrote:I will review all the articles. But just thinking out loud since someone does not know if growth or value will occur in the future, this would be the reason for blended fund. Instead of large growth large blend etc. Looks this way when backtesting. Or even better the total fund for the sector. If that makes sense.


Theory suggests that the most efficient portfolio is the market portfolio. Ignoring costs, one wants to hold the whole market - large and small, foreign and domestic.

However, I am not sure if you are asking the right question . Different factors act differently. This ties into your S&P 500 v. Total market thread. Value offers both higher returns and higher risks. Low Beta offers lower returns and lower risks. Which fits your goals, required returns, and risk tolerances?

This is a technical paper that goes into greater detail.
http://www.cfapubs.org/doi/abs/10.2469/faj.v72.n6.3

indexonlyplease
Posts: 526
Joined: Thu Apr 30, 2015 12:30 pm
Location: Pembroke Pines, FL

Re: backtesting different indexes

Postby indexonlyplease » Tue Jul 18, 2017 10:27 am

alex_686 wrote:
indexonlyplease wrote:I will review all the articles. But just thinking out loud since someone does not know if growth or value will occur in the future, this would be the reason for blended fund. Instead of large growth large blend etc. Looks this way when backtesting. Or even better the total fund for the sector. If that makes sense.


Theory suggests that the most efficient portfolio is the market portfolio. Ignoring costs, one wants to hold the whole market - large and small, foreign and domestic.

However, I am not sure if you are asking the right question . Different factors act differently. This ties into your S&P 500 v. Total market thread. Value offers both higher returns and higher risks. Low Beta offers lower returns and lower risks. Which fits your goals, required returns, and risk tolerances?

This is a technical paper that goes into greater detail.
http://www.cfapubs.org/doi/abs/10.2469/faj.v72.n6.3


At this time in my life you are right that is the big question. risk vs returns. I am at the point that I can say I have enough with a great pension and great investment account.. A change in the market would not mean much to me since I already have income and don't rely on investment income. So, since I am happy with my 3 fund. I am at a time of more understanding which can be scary. Why can I say that, becasue when I knew a lot less years ago, I did less with my investments. I just funded the 457 plan maxed out when I could in mutual funds I thought were good. I went through dot.com, 2008 and did nothing but the same funding the 457. Mid 40 found out about Roth IRA and maxed that out to. Not that I knew anything I just did not know any different and was buy working 3 jobs, all jobs I enjoyed. I guess that kept me away from looking at the accounts. Also, not sure when the account started on the computer but I remember getting the quarterly statements and just throwing them in the garbage and never open. Why did not know what I was looking at.

I think my learning is now to help my kids in investing education.

Thanks for the continue help in question and more to come.

cliff


Return to “Investing - Help with Personal Investments”

Who is online

Users browsing this forum: BanquetBeer, bob_m10, Born2Run, EinAustin, merits, MoonOrb, RadAudit and 77 guests