2015 portfolio performance

[Jurgis]

Using the XIRR function, you get the compounded return while depositing and withdrawing cash in between.

 

How would you compare against a benchmark return similarly, i.e. DCA / monthly payments into S&P, instead of point to point return?

 

That's also relevant observation. Unfortunately, in cases like mine, I can't really compare to "similar" returns from index. I just compare to point-to-point - yearly - index returns.

[Hielko]

I wonder why others don't simply use the XIRR function, given how easy it is to do. I wonder how much internet stock performance is rather inaccurate because of this.

Using IRR is not necessarily the correct way to calculate performance. You can also make a good case for calculating a time-weighted rate of return, but that is usually pretty hard to calculate if your broker isn't doing that for you. I use an IRR calculation because that is the only practical option when you have to combine results from multiple brokers.

[randomep]

I wonder why others don't simply use the XIRR function, given how easy it is to do. I wonder how much internet stock performance is rather inaccurate because of this.

 

 

 

+1

 

I was just about to say the same thing, I think few people managing their own finances know abuot it. My life changed completely when someone on CoBF brought it to my attention. okok maybe a slight exaggeration.....

[EliG]

Bogleheads spreadsheet calculates both dollar-weighted and time-weighted returns.

 

https://www.bogleheads.org/wiki/Calculating_personal_returns

 

 

[randomep]

I wonder why others don't simply use the XIRR function, given how easy it is to do. I wonder how much internet stock performance is rather inaccurate because of this.

Using IRR is not necessarily the correct way to calculate performance. You can also make a good case for calculating a time-weighted rate of return, but that is usually pretty hard to calculate if your broker isn't doing that for you. I use an IRR calculation because that is the only practical option when you have to combine results from multiple brokers.

 

Hielko, what do you mean time-weighted return?

 

irr is just the fixed rate of return for all contributions until now, which will give you the current account balance that is marked to market. That's how mutual funds post their trailed 1yr, 5yr, 10yr and from-inception returns......

 

 

[Hielko]

I wonder why others don't simply use the XIRR function, given how easy it is to do. I wonder how much internet stock performance is rather inaccurate because of this.

Using IRR is not necessarily the correct way to calculate performance. You can also make a good case for calculating a time-weighted rate of return, but that is usually pretty hard to calculate if your broker isn't doing that for you. I use an IRR calculation because that is the only practical option when you have to combine results from multiple brokers.

 

Hielko, what do you mean time-weighted return?

 

irr is just the fixed rate of return for all contributions until now, which will give you the current account balance that is marked to market. That's how mutual funds post their trailed 1yr, 5yr, 10yr and from-inception returns......

No: mutual funds calculate time-weighted returns, not internal rates of return. When you calculate a time-weighted return you eliminate the impact of deposits and withdrawals which might be appropriate since you might have no or a limited influence on when you make them (for mutual funds it's determined by their investors while for you personally it might be determined by when you receive your salary). If you try to time the market by deciding when to add cash you should use an IRR.

 

Use google to check the difference...

[Guest Grey512]

Looking at the bell-curve from the votes, I wonder whether the average return of the COBF community is better or worse than that of the hedge fund community (e.g. the HFRI index?). Looks a bit better.

[benhacker]

The best way to think about IRR vs. TWRR is this, IMO:

 

1) IRR takes all inflows and outflows into account, and provides the effective (annual) rate of return which equates the present and future value of the cash flows (the effective annual rate).

 

2) TWRR is a cash flow neutral measure, so it weights the return of the underlying daily investment results equally, even if more or less money was invested on certain days of the measurement period.

 

If you think about this, IRR is best used for the following:

1) Investors who control fully the inflows and outflows of money (I would argue, for a personal investor, with knowledge of what IRR means, this is the appropriate measure).

 

2) If you are an Advisor**, or HF or something and you don't control the timing of when *investors* give you money, TWRR is more appropriate measure as you want to separate manager skill vs. investors good / bad timing.  There is a gray area here as perhaps some investors do indeed influence the timing of when their clients give them money.

 

** Note that some (most?) PE firms lock investor money into a strategy based on commitments, but don't measure their (IRR) returns until they "call" the money into the strategy.  This effectively makes PE returns seem higher, but I think is fundamentally an incorrect way to measure returns as "locked" in money should be considered committed to the strategy.

 

TWRR is very laborious to calculate.

 

Regardless of the above, it should be noted that that IRR can't be used and compared directly to the returns of an index, because you don't normalize the timing of cashflows into that index in the same way (at least 99% of people don't)... if you have a very large account, and your cash inflows (or outflows) are small relative to the starting / ending value of the account(s) you are measuring, then it doesn't likely matter much.  If your cash flows into your account are large as % of starting value, your IRR will be much more volatility in general around whatever return average you would effectively have reached.

 

Hope my words conveyed the nuance here...

 

I quote both returns to my clients, and I say basically that IRR is *their* return, and TWRR is *my* return... basically the best way to simplify it I have found...

[SharperDingaan]

-19% TWRR

 

Primarily from hedging too late on PWT. Absent PWT & we're more or less flat with fixed income gains offsetting equity losses. Not a big deal for us, as we withdrew all of our original capital long ago - & this is all house money. We are accumulating at rock bottom prices, & should do very well 5 years out.

 

Were we not expecting such ridiculous pricing over the next 1-2 years, we would be in floating rate FI & targeting an overall 8%+ cash yield. But if you have the structure, & can tolerate the risk, it is pretty hard to see why you would not have significant weightings to the  Glencore's/Plats/Anglo Americans of the world right now. Again, the long term view.

 

SD

 

[innerscorecard]

I know that in the investment industry, you are supposed to use time-weighted returns. This is to easily compare returns between different funds.

 

I like using dollar-weighted returns, because I do think the timing of deploying capital is a measure of investment skill. But I benchmark my annualized dollar-weighted returns to a simple annual return for the relevant index ETFs, because if I were passively indexing, the cash deployed would be regular.

[randomep]

I wonder why others don't simply use the XIRR function, given how easy it is to do. I wonder how much internet stock performance is rather inaccurate because of this.

Using IRR is not necessarily the correct way to calculate performance. You can also make a good case for calculating a time-weighted rate of return, but that is usually pretty hard to calculate if your broker isn't doing that for you. I use an IRR calculation because that is the only practical option when you have to combine results from multiple brokers.

 

Hielko, what do you mean time-weighted return?

 

irr is just the fixed rate of return for all contributions until now, which will give you the current account balance that is marked to market. That's how mutual funds post their trailed 1yr, 5yr, 10yr and from-inception returns......

No: mutual funds calculate time-weighted returns, not internal rates of return. When you calculate a time-weighted return you eliminate the impact of deposits and withdrawals which might be appropriate since you might have no or a limited influence on when you make them (for mutual funds it's determined by their investors while for you personally it might be determined by when you receive your salary). If you try to time the market by deciding when to add cash you should use an IRR.

 

Use google to check the difference...

 

Ok I can't google any place that gives the formula for mutual fund returns but I'll pose some scenarios. Folks please feel free to chime in.

 

Suppose Brk is at $1000 in 2011, by 2016 it is $2000. So its IRR or whatever you want to call it is about 15% right? They pay no dividends.

 

Suppose you have a mutual fund that owns nothing but brk, then its advertized rate of return over those five years is also 15%?

 

Now suppose you have a fund that pays dividends and capital gain distributions annually. Then isn't the return the irr calculated from the initial NAV, the annual distributions and the final NAV?

 

So for example for one fund I would enter on the spreadsheet

 

-365 (initial NAV)

10 (distribution 1)

10 (distribution 2)

10 (distribution 3)

10 (distribution 4)

495 (final price + distribution 5)

 

so the advertised 5yr return is the irr of the above 6 cells?

 

 

 

 

[Hielko]

I wonder why others don't simply use the XIRR function, given how easy it is to do. I wonder how much internet stock performance is rather inaccurate because of this.

Using IRR is not necessarily the correct way to calculate performance. You can also make a good case for calculating a time-weighted rate of return, but that is usually pretty hard to calculate if your broker isn't doing that for you. I use an IRR calculation because that is the only practical option when you have to combine results from multiple brokers.

 

Hielko, what do you mean time-weighted return?

 

irr is just the fixed rate of return for all contributions until now, which will give you the current account balance that is marked to market. That's how mutual funds post their trailed 1yr, 5yr, 10yr and from-inception returns......

No: mutual funds calculate time-weighted returns, not internal rates of return. When you calculate a time-weighted return you eliminate the impact of deposits and withdrawals which might be appropriate since you might have no or a limited influence on when you make them (for mutual funds it's determined by their investors while for you personally it might be determined by when you receive your salary). If you try to time the market by deciding when to add cash you should use an IRR.

 

Use google to check the difference...

 

Ok I can't google any place that gives the formula for mutual fund returns but I'll pose some scenarios. Folks please feel free to chime in.

 

Suppose Brk is at $1000 in 2011, by 2016 it is $2000. So its IRR or whatever you want to call it is about 15% right? They pay no dividends.

 

Suppose you have a mutual fund that owns nothing but brk, then its advertized rate of return over those five years is also 15%?

 

Now suppose you have a fund that pays dividends and capital gain distributions annually. Then isn't the return the irr calculated from the initial NAV, the annual distributions and the final NAV?

 

So for example for one fund I would enter on the spreadsheet

 

-365 (initial NAV)

10 (distribution 1)

10 (distribution 2)

10 (distribution 3)

10 (distribution 4)

495 (final price + distribution 5)

 

so the advertised 5yr return is the irr of the above 6 cells?

http://lmgtfy.com/?q=time+weighted+return+calculation

[Guest JoelS]

Portfolio roughly flat in USD at year end.

Measured in home currency, IRR of approximately 15% for the period.

 

As regards IRR, Howard Marks wrote a good memo on the subject in 2006: https://www.oaktreecapital.com/docs/default-source/memos/2006-07-12-you-cant-eat-irr.pdf?sfvrsn=2

 

 

[investor-man]

Portfolio roughly flat in USD at year end.

Measured in home currency, IRR of approximately 15% for the period.

 

As regards IRR, Howard Marks wrote a good memo on the subject in 2006: https://www.oaktreecapital.com/docs/default-source/memos/2006-07-12-you-cant-eat-irr.pdf?sfvrsn=2

 

That's a great memo. And I think IRR is the correct metric by which to judge the performance of someone like an individual investor, who is completely in control of his/her own capital. IRR, or something that's abstractly similar (ROIC, ROE, etc.), is what we all use to judge all of these companies we're investing in. Why wouldn't we use it on ourselves? The time weighted metric just seems like a way to avoid the truth.

[rohitc99]

Looking at the bell-curve from the votes, I wonder whether the average return of the COBF community is better or worse than that of the hedge fund community (e.g. the HFRI index?). Looks a bit better.

 

different investing styles, different currencies and ofcourse different countries if not continents. are these results representative of anything ? on individual basis maybe, but we dont know what is getting aggregated in the graph

[Hielko]

Portfolio roughly flat in USD at year end.

Measured in home currency, IRR of approximately 15% for the period.

 

As regards IRR, Howard Marks wrote a good memo on the subject in 2006: https://www.oaktreecapital.com/docs/default-source/memos/2006-07-12-you-cant-eat-irr.pdf?sfvrsn=2

 

That's a great memo. And I think IRR is the correct metric by which to judge the performance of someone like an individual investor, who is completely in control of his/her own capital. IRR, or something that's abstractly similar (ROIC, ROE, etc.), is what we all use to judge all of these companies we're investing in. Why wouldn't we use it on ourselves? The time weighted metric just seems like a way to avoid the truth.

Most individual investors are absolutely not fully control of their own capital. You can't change when you receive your salary, or when you get an inheritance or a gift etc. If you got a big gift in 2009 you just got lucky, while if you got it in 2007 you got unlucky. If you don't try to time the market does it make sense to try to measure that?

[kab60]

Thanks for the input re calculating portfolio returns. 2015 was my first full year of investing and I was up 12,5 percent. I calculated my return comparing the total amount of capital in my brokerage account at yearend with the total amount of capital I had transferred to my brokerage account. I committed most of my capital throughout the year so IRR is probably higher but it doesn't really matter to me. I know how much I set aside for investing and how much it grew and that'll do.

[Uccmal]

Thanks for the input re calculating portfolio returns. 2015 was my first full year of investing and I was up 12,5 percent. I calculated my return comparing the total amount of capital in my brokerage account at yearend with the total amount of capital I had transferred to my brokerage account. I committed most of my capital throughout the year so IRR is probably higher but it doesn't really matter to me. I know how much I set aside for investing and how much it grew and that'll do.

 

Makes sense to me.  Good result BTW. 

 

Over time the differences in the way things are calculated become less pronounced.  For my total average cagr I just took the amount from 11 years ago and added back the net withdrawals.  It gives me an estimate +/- a couple percent, nothing more. If its lower than 22%, say 18%, then it still indicates that I beat the returns of all mutual funds, and all etfs, over the same 11 year period. 

 

And we have spent so much money that I dont need to worry about size becoming an anchor any time soon.  My family is very good at taking my high yielding assets, and reinvesting the proceeds in low yielding assets (home/cottage).  "if I was single my pockets would jingle.....".

[oddballstocks]

Portfolio roughly flat in USD at year end.

Measured in home currency, IRR of approximately 15% for the period.

 

As regards IRR, Howard Marks wrote a good memo on the subject in 2006: https://www.oaktreecapital.com/docs/default-source/memos/2006-07-12-you-cant-eat-irr.pdf?sfvrsn=2

 

That's a great memo. And I think IRR is the correct metric by which to judge the performance of someone like an individual investor, who is completely in control of his/her own capital. IRR, or something that's abstractly similar (ROIC, ROE, etc.), is what we all use to judge all of these companies we're investing in. Why wouldn't we use it on ourselves? The time weighted metric just seems like a way to avoid the truth.

Most individual investors are absolutely not fully control of their own capital. You can't change when you receive your salary, or when you get an inheritance or a gift etc. If you got a big gift in 2009 you just got lucky, while if you got it in 2007 you got unlucky. If you don't try to time the market does it make sense to try to measure that?

 

Agree with this, that is unless people are literally hoarding money, storing it away without ever intending to use it that timing is really out of our control.

 

I have no idea how I did in 2015, calculating it will be a mess because money was moving in and out throughout the year.  We purchased a house halfway through the year and I sold some stocks to facilitate the purchase.  Would I have done better if I hadn't sold them? Maybe, but then again I wouldn't be living where I want to.  In my view the purpose of money is to work for you, and in this case investing enabled us to buy a place we wanted.

 

I'd like to think I can control the timing of my account, but I can't.  When we found a house I couldn't wait until the year end to make performance nice and tidy.

[cwericb]

-1.27%

Very disappointing considering I was up by 7% in April. Since the Dow was flat and Toronto was down about 7% in 2015, I guess I shouldn’t complain. But after being up 16% in 2014 it’s a bit of a kick in the teeth.

 

I track my results separate from an account I have invested through a financial guy and for the first time ever he beat my results, primarily through currency, which gave me an overall result of +1.2%.

 

I’ve done worse.  :)

 

Oddballstocks: Can’t go wrong with the home purchase. Markets could nosedive, housing could dip, but you still have a roof over your head and the home in which you chose to live.

[Liberty]

Did about 20% 2015, but I'm not that happy because FX helped a lot and without it, results would be a fair bit worse.

 

I feel like 2015 was a particularly exhausting year for most market participant. That's good. I prefer everyone to be mistrustful and worn out than happy and contented.

[Uccmal]

Portfolio roughly flat in USD at year end.

Measured in home currency, IRR of approximately 15% for the period.

 

As regards IRR, Howard Marks wrote a good memo on the subject in 2006: https://www.oaktreecapital.com/docs/default-source/memos/2006-07-12-you-cant-eat-irr.pdf?sfvrsn=2

 

That's a great memo. And I think IRR is the correct metric by which to judge the performance of someone like an individual investor, who is completely in control of his/her own capital. IRR, or something that's abstractly similar (ROIC, ROE, etc.), is what we all use to judge all of these companies we're investing in. Why wouldn't we use it on ourselves? The time weighted metric just seems like a way to avoid the truth.

Most individual investors are absolutely not fully control of their own capital. You can't change when you receive your salary, or when you get an inheritance or a gift etc. If you got a big gift in 2009 you just got lucky, while if you got it in 2007 you got unlucky. If you don't try to time the market does it make sense to try to measure that?

 

Agree with this, that is unless people are literally hoarding money, storing it away without ever intending to use it that timing is really out of our control.

 

I have no idea how I did in 2015, calculating it will be a mess because money was moving in and out throughout the year.  We purchased a house halfway through the year and I sold some stocks to facilitate the purchase.  Would I have done better if I hadn't sold them? Maybe, but then again I wouldn't be living where I want to.  In my view the purpose of money is to work for you, and in this case investing enabled us to buy a place we wanted.

 

I'd like to think I can control the timing of my account, but I can't.  When we found a house I couldn't wait until the year end to make performance nice and tidy.

 

We bought a cottage.  My total losses could have been even worse.  I probably would have lost the value of what we spent using the same percentages.  Not whining about my first world problems, thats for sure.  Stocks will come back.

[zarley]

-9% +/- in my personal accounts.  Up a percent or two in indexed retirement accounts.

 

Easily my worst year since 2008, mostly due to a lot of BRK and little time devoted to managing my accounts.  Not terribly worried about it given that BRK is undervalued and I'm not graded by anyone but myself on the year to year performance.  I may have missed a few opportunities due to the lack of time/effort on my part, but that doesn't bother me a lot either.

[Jurgis]

Re Uccmal and oddball - buying house at the market top is usually great thing. Not counting 2007 of course. I bought my first house in 2000 - great timing to sell some stocks ready for Internet bubble burst. Not that I did it on purpose. It's great to be lucky. ;)

 

 

[investor-man]

Portfolio roughly flat in USD at year end.

Measured in home currency, IRR of approximately 15% for the period.

 

As regards IRR, Howard Marks wrote a good memo on the subject in 2006: https://www.oaktreecapital.com/docs/default-source/memos/2006-07-12-you-cant-eat-irr.pdf?sfvrsn=2

 

That's a great memo. And I think IRR is the correct metric by which to judge the performance of someone like an individual investor, who is completely in control of his/her own capital. IRR, or something that's abstractly similar (ROIC, ROE, etc.), is what we all use to judge all of these companies we're investing in. Why wouldn't we use it on ourselves? The time weighted metric just seems like a way to avoid the truth.

Most individual investors are absolutely not fully control of their own capital. You can't change when you receive your salary, or when you get an inheritance or a gift etc. If you got a big gift in 2009 you just got lucky, while if you got it in 2007 you got unlucky. If you don't try to time the market does it make sense to try to measure that?

 

Agree with this, that is unless people are literally hoarding money, storing it away without ever intending to use it that timing is really out of our control.

 

I have no idea how I did in 2015, calculating it will be a mess because money was moving in and out throughout the year.  We purchased a house halfway through the year and I sold some stocks to facilitate the purchase.  Would I have done better if I hadn't sold them? Maybe, but then again I wouldn't be living where I want to.  In my view the purpose of money is to work for you, and in this case investing enabled us to buy a place we wanted.

 

I'd like to think I can control the timing of my account, but I can't.  When we found a house I couldn't wait until the year end to make performance nice and tidy.

 

The scenario you describe above is precisely why XIRR is the good metric to use when evaluating your returns: you'd like the return that accounts for the large cashflow out of your account(s).

 

I think people are making this calculation more difficult than it needs to be. You basically need a spreadsheet with two columns: dates and cashflows. For example, I have 6 accounts between my wife and I, but I don't keep track of what's happening in each separately. I just keep track of when deposits and withdrawals are made. Which account the transaction occurred in is unimportant. I had roughly 30 such transactions during 2015, and keeping track of that is not laborious in any way. Are people making 100s/1000s of transactions in and out of their trading accounts (keep in mind if you make 100s/1000s of transactions WITHIN your account, XIRR accounts for this and you don't need to keep track of it)? I could see how that is laborious, but I can't imagine a reason why someone would do this.

 

And also...

 

In my description of why XIRR is a good metric for individual investors, I did say they have "complete control". Ok you got me, they don't have "complete control", but nobody using this metric in a real-life scenario has "complete control" over any of the cash-flows they are experiencing, and it's never assumed to be the case. That said, an individual has pretty reliable control over their investments: they are typically receiving a steady cashflow of income that they can choose to invest at their discretion.  And I don't understand the comment about getting a gift in 2007 vs 2009 and the "luck" or timing involved. Yeah it would have been better to receive it in 2009 instead of 2007, but that is irrelevant. Don't you simply want to measure performance? Maybe you're making the argument that a good investment made in December will push my XIRR for the year up higher than a good investment made in January even if I end the year at the same dollar amount? This is mathematically correct. That said, you should keep track of your XIRR for your entire investment lifetime, so you get an accurate number reflecting your overall performance.

 

Ok, I'm done :)