(Mis?) comparing your performance vs an index

[Golden Mean]

Finance YouTuber Matt Derron posted an interesting video about how his Schwab? account incorrectly compares his portfolio performance to the S&P 500. This same issue appears to be the case with the portfolio software I use. (They also miscalculated his actual gains too but that's an issue with a lost of cost basis on a initial transfer he did.)

The issue is: If you are buying equities, and your software charts your performance, it is wrong to compare your percentage gain vs the straight percentage gain for whatever index you want to compare to during that same time frame.

For example, if you bought an equity in Feb and again in May, your typical portfolio software is going to graph/compare those equity's percent change vs the S&P 500 from Feb to today. The problem is you need the S&P 500 percent change to be calculated using (simulated/calculated) purchases of the S&P 500 on same dates as the equities you purchased. If you don't, you are comparing your perf vs buying the S&P 500 with all of your capital on the first date.

I put a spreadsheet together that I hope shows the issue. It compares VOO to itself. You can see that two purchases of VOO through the time frame cannot be compared to the percent change of VOO throughout that same time frame.

Hope someone finds this enlightening or shows me how I've got the wrong end of the stick.

 

[Golden Mean]

Here's an example of the type of "compare your performance to the market" graphs I'm referring to. It shows the S&P 500 up about 34%, but I'd actually be up more like 25% if I had bought the index on same dates as I bought stock and with the same funds.

 

[JOBO]

I started using the same website he does a few months ago and noticed the same thing.
Additionally, indices don’t have trading costs, cash drag, redemptions or taxes so real investor returns will always be less than published index returns.
My portfolio hasnt been tightly bound to any one index since ~2010 so I never really thought about it but I've long held that I dont need to beat the market, my brother or any other benchmark to meet my goals.

 

[Al18]

Most stocks are overpriced right now. Warren Buffet has been a net seller of stocks in the past 6 months. I’m not buying until I see a 10% sale.

 

[SecondCor521]

OP, I didn't check your specific examples, but generally you've got the right idea.

What you want to use in all cases is IRR, which can be done in Excel with the XIRR() function. You just need to dates and amounts of all cash inflows and outflows, plus the ending balance.

If you're comparing against an index, you also probably want to account for the dividends and interest paid by the stocks in the index. Most indexes provided by most sites just provide the price change only, which undercounts the index's performance.

I also think people should always account for risk and taxes in this analysis, and do so over a reasonably long period of time across all of their investments. But in the last 20 years or so, I've not met a single person who does all of those things and does them properly. (I don't do it properly either, I just mostly VTSAX and chill.)

FWIW, Vanguard uses IRR when they inform you of your portfolio performance.

 

[PlayinwithFIRE]

You are talking about cost basis I think. Not a new concept and you'd think it would be part and parcel to any calculations a website might make to properly estimate gains/losses.

What's the point of all that computing power and data mining if it isn't invoked and its left to us pen and paper guys to track/estimate it?

 

[jazz4cash]

I thought the youtuber had a very basic unforgivable misunderstanding of SP500 performance reports. Shot himself in the foot. I don’t understand why he thought the index results were based on the timing of his contributions.

 

[Golden Mean]

You are talking about cost basis I think. Not a new concept and you'd think it would be part and parcel to any calculations a website might make to properly estimate gains/losses.

Yes, I think you are right about it being about essentially calculating your cost basis on the index instead of just checking the index's change from a start to an end date.

What's the point of all that computing power and data mining if it isn't invoked and its left to us pen and paper guys to track/estimate it?

I agree, but the site I use does it "wrong" and they really didn't seem that concerned when I pointed it out

To me the whole point of tracking my performance with individual stocks is so I can see if I'm actually doing better buying needles in the haystack over buying the whole haystack.

I thought the youtuber had a very basic unforgivable misunderstanding of SP500 performance reports. Shot himself in the foot. I don’t understand why he thought the index results were based on the timing of his contributions.

My guess would be that the S&P 500 percentage change is presented in a way that suggests you compare that number to the numbers for your performance. ie it's on the same page/table/chart.

 

[Golden Mean]

...
If you're comparing against an index, you also probably want to account for the dividends and interest paid by the stocks in the index. Most indexes provided by most sites just provide the price change only, which undercounts the index's performance.

You are right, I should be calculating the index return with dividends. I'll look into it and see how hard it'd be to fold into my spreadsheet.

...

FWIW, Vanguard uses IRR when they inform you of your portfolio performance.

Yeah, they do, but I don't see anyway that they let you compare that number to an index.

 

[RobLJ]

As a quick comparison, I would compare my portfolio performance to the Fidelity retirement fund that most closely matched my stock/bond allocation. This doesn't deal with the issue of buying during the year; if you use total gain that (largely) deals with the dividend issue.
As long as the portfolio performance was relatively close to the comparable retirement fund, I felt fine (usually was, but the international allocation and other quirks in my portfolio did lead to some years of over and underperformance.)

 

[Golden Mean]

As a quick comparison, I would compare my portfolio performance to the Fidelity retirement fund that most closely matched my stock/bond allocation. This doesn't deal with the issue of buying during the year; if you use total gain that (largely) deals with the dividend issue.

Total gain does help with the dividend issue (unfortunately I can't get that for idx funds/etfs from the GOOGELFINANCE function in Google Sheets), but I'd argue you can't get useful info from from the comparison you are doing. Say you started investing in Feb. 2019 and put $50k into a fund/stocks periodically over the next 5 years. Today, if you compare your gain to just the percent change of that retirement fund since 2019, you'd be comparing 5 years of buys at various prices vs having bought $50k of that retirement fund in Feb. of 2019.

 

[Dalton]

Most stocks are overpriced right now. Warren Buffet has been a net seller of stocks in the past 6 months. I’m not buying until I see a 10% sale.

I said the same thing a year ago. I'm thankful, now, that I didn't stick to my guns.

As to the OP, it's all relative. If you look at the S&P or Total US Stock as the "benchmark", you can get an idea of where you are, relatively. It's just important to understand that trying to make a direct comparison isn't appropriate. Also, important to remember that the smaller your portfolio is, the greater impact new money that was added over that time frame will affect any comparison (e.g. adding $20k to a $100k portfolio has a much greater impact than adding the same $20k to a $1 million portfolio).

 

[Lsbcal]

What I do is to create 2 portfolios that I might consider as alternatives. Also a minimum period for comparing performance should be a few years I think. Since I bounce around in index funds within tax advantaged accounts my alternative portfolios to compare are (1) a Wellington based portfolio so I do not have to make changes in fund types, (2) a fixed SP500 portfolio which includes fixed international as well.

Well some here will say that this is way overkill but what else should I do with my retirement time
So here is a spreadsheet snippet of the comparison:

Comparing the red circled summaries the performance of 1.130 in my portfolio (read as 13.0%) is better then the other 2 portfolios for the period from March 1993 to today. Likewise for other periods it does better. Since January 2023 my actual portfolio is about 2.4% annually better then the other two. For a largish portfolio this translates into lots of extra spending dollars.

Now onto buying those rugs and chairs DW wants.

 

[Golden Mean]

What I do is to create 2 portfolios that I might consider as alternatives. Also a minimum period for comparing performance should be a few years I think. Since I bounce around in index funds within tax advantaged accounts my alternative portfolios to compare are (1) a Wellington based portfolio so I do not have to make changes in fund types, (2) a fixed SP500 portfolio which includes fixed international as well.

Well some here will say that this is way overkill but what else should I do with my retirement time
So here is a spreadsheet snippet of the comparison:

View attachment 52707

Comparing the red circled summaries the performance of 1.130 in my portfolio (read as 13.0%) is better then the other 2 portfolios for the period from March 1993 to today. Likewise for other periods it does better. Since January 2023 my actual portfolio is about 2.4% annually better then the other two. For a largish portfolio this translates into lots of extra spending dollars.

Now onto buying those rugs and chairs DW wants.

Nice! Spreadsheets rule.

What I've done is created a spreadsheet where I enter my stock trades (I don't analyze my index holdings) and have the googlefinance() function look up the price of the index on those same dates, so I get a direct comparison. It's enlightening.

Here's a trade I thought I got out of "OK", but clearly would have been better in the index fund:

 

[SecondCor521]

^ It isn't a direct comparison because you're ignoring income events with both.

 

[RobLJ]

Total gain does help with the dividend issue (unfortunately I can't get that for idx funds/etfs from the GOOGELFINANCE function in Google Sheets), but I'd argue you can't get useful info from from the comparison you are doing. Say you started investing in Feb. 2019 and put $50k into a fund/stocks periodically over the next 5 years. Today, if you compare your gain to just the percent change of that retirement fund since 2019, you'd be comparing 5 years of buys at various prices vs having bought $50k of that retirement fund in Feb. of 2019.

You are exactly right, although I found as the portfolio got larger, the effects of incremental buys had less and less of an effect. If you are in the first 15 years or so of investing, what you are describing is important, but I admit it didn't seem worth the time or effort to me, but to many here it would matter a lot. (Not knocking it.) My view was that close enough was good enough. I had two kids, and Aged Ps and a degenerating twin with them, so my time seemed valuable. Now I have the time but don't care, but that seems pretty flippant even to me.

 

[Golden Mean]

^ It isn't a direct comparison because you're ignoring income events with both.

Are you referring to dividends? I am tracking those in my spreadsheet now for my current holdings but I haven't found a way to get them for VOO in an automated way. Looks like the S&P 500's yield hovers between 1.3% and 2% most of the time.

 

[Golden Mean]

You are exactly right, although I found as the portfolio got larger, the effects of incremental buys had less and less of an effect. If you are in the first 15 years or so of investing, what you are describing is important, but I admit it didn't seem worth the time or effort to me, but to many here it would matter a lot. (Not knocking it.) My view was that close enough was good enough. I had two kids, and Aged Ps and a degenerating twin with them, so my time seemed valuable. Now I have the time but don't care, but that seems pretty flippant even to me.

Oh yeah, makes sense to me. I only invested in index funds during my working career. Now that I'm retired I've got 10% of my portfolio in individual stocks and want to make sure I'm not fooling myself that I'm doing better than just holding an index fund. i.e. It's a hobby

 

[SecondCor521]

Are you referring to dividends? I am tracking those in my spreadsheet now for my current holdings but I haven't found a way to get them for VOO in an automated way. Looks like the S&P 500's yield hovers between 1.3% and 2% most of the time.

Income events would be dividends, interest, and capital gains distributions.

They could also include stock spinoffs, but those are rare, relatively speaking.

If you're trading frequently, I think a person should also adjust for taxes and risk.

 

[Golden Mean]

Income events would be dividends, interest, and capital gains distributions.

Got it. I'm not aware of VOO doing any capital gains dists or paying interest.

They could also include stock spinoffs, but those are rare, relatively speaking.

Are you referring to an S&P 500 index fund doing these?

If you're trading frequently, I think a person should also adjust for taxes and risk.

True. I do my stock work in an IRA to keep it simple.

 

[Lsbcal]

VOO has quarterly dividends. Click on distributions here:
VOO page

 

[Golden Mean]

VOO has quarterly dividends. Click on distributions here:
VOO page

Right, which I mentioned in post #17, but it doesn't have any capital gains dists or interest payments afaik.

 

[SecondCor521]

Are you referring to an S&P 500 index fund doing these?

Not specifically; it was just a side comment in the context of talking about income events and comparing performance.

I haven't looked lately, but I would guess that stock spinoffs from companies in the S&P500 would be uncommon to rare. When they happen, that would reduce the capitalization of the companies involved, which would change the index. I'm sure they adjust the index automatically to account for such spinoffs. I would also guess that the companies spun off are typically small enough to where they are not big enough to be in the S&P500.

All in all, spinoffs are mostly a technicality in the context of the S&P500. They're more interesting (and more hassle IMHO) when they happen to stocks that you own individually in taxable accounts because the basis tracking gets tricky. Or at least it used to before basis became mostly tracked by the brokerages.