Do you net out investment expenses from your SWR?

[ESRBob]

I thought I had learned that the academid SWR analyses (e.g. Trinity Study, Bengen) all used raw market data, with no investment management fees involved. That would suggest the need to pay for those fees out of the SWR. Do others agree? How do yo do your calculations?

In other words, if you are basing things around a 4% SWR, and you have average fees of 0.4% and estimated trading costs (even in a TM or index fund) of another 0.1% or 0.2%, should we in fact be only withdrawing for spending the net of those, or something along the lines of 3.4% or 3.5%?

I know it sounds depressing, but that is what I thought I understood from my research -- anyone else doing the same calculations?

ESRBob

 

[cute fuzzy bunny]

Re: Do you net out investment expenses from your S

My aggregate management fees are roughly .22%, small enough to not worry much about factoring in.

You can get these below .20 easily, and at .15 if your primary holding is vangard admiral indexes.

Mine only creep up because I own a couple of international funds, the explorer fund, the healthcare fund, etc that are a little higher. Not high enough to make me want to abandon some asset allocation though.

 

[malakito]

Re: Do you net out investment expenses from your S

Good question, a couple of answers for you:

1. You're right, fees should be subtracted from the 4% as another cost of living. This is a great reason to get them down to 10-20 basis points. (Personally I'm around 15 basis points and dropping.)

2. There have been articles -- sorry, can't find the exact cite for you now -- that show that Vanguard has historically been able, with some strategies that they've developed, to beat the index they track by somewhere around 9 basis points, give or take. This essentially cuts their fees in half.

3. If any of your planning calculations are done based on the returns from any of your actual mutual funds, remember that the performance reported is usually net of management fees, so you shouldn't have to subtract it again.

Personally I have a little spreadsheet that calculates out these fees for me for each account I have and then does an average on the whole nut. It helps me see where I can look to cut costs -- for example we have some funds in VPMCX, which has an expense ratio of 51 basis points. Low, but above average, so I think about asking the wife about switching that to VFINX or something.

malakito.

 

[Cut-Throat]

Re: Do you net out investment expenses from your S

FIRECalc lets you plug any number you want into it. That's about as accurate as I need to get.

 

[unclemick]

Re: Do you net out investment expenses from your S

Bernstein has visited tracking error several times - go to Efficient Frontier, put in index tracking error on his site search. Being a boglehead - I like sept. 98.

 

[Bob_Smith]

Re: Do you net out investment expenses from your S

This is another small advantage of owning individual TIPS. It lowers my overall expenses. I'm somewhere in the vicinity of 0.15% on the total (plus those hidden trading costs my funds encounter and that are NOT included in the expense ratio, as ESRBob mentions).

I'm also mindful of the fact that Vanguard does screw-up on occasion and significantly under-perform the index they're tracking. For example, in 2002 their Total Bond Market Index Fund underperformed the Lehman Aggregate Bond Index significantly (Vanguard @ 8.26% vs. Index @ 10.26%). That error cost me several thousand bucks.

 

[cute fuzzy bunny]

Re: Do you net out investment expenses from your S

Yeah they've done that quite a few times with several funds.

The wellesley fund underperformed the concocted "wellesley index" for the past six months.

Thats what makes me feel squeamish about their recent changes in the bond fund names to allow them more leeway in actively managing these funds if they so choose.

Much more of this fiddling and I may look to another fund company. When I buy a damn index I want that index. For a company that espouses the power of indexing and have brought in billions on the back of that, they appear quite willing to screw that up.

 

[intercst]

Re: Do you net out investment expenses from your S

I thought I had learned that the academid SWR analyses (e.g. Trinity Study, Bengen) all used raw market data, with no investment management fees involved.  That would suggest the need to pay for those fees out of the SWR.  Do others agree?  How do yo do your calculations?

In other words, if you are basing things around a 4% SWR, and you have average fees of 0.4% and estimated trading costs (even in a TM or index fund) of another 0.1% or 0.2%, should we in fact be only withdrawing for spending the net of those, or something along the lines of 3.4% or 3.5%?

I know it sounds depressing, but that is what I thought I understood from my research -- anyone else doing the same calculations?

ESRBob

That's correct. Bengen, Trinity Study, etc. ignore trading costs and mutual fund fees.

Both FIRECalc and the REHP Excel spreadsheet allow you to input your investment expenses into the calculation.

As a minimum you should use the expense ratio for the fund (e.g. 0.18% annual expense ratio for the Vanguard S&P500 index fund.)

If you want to get really accurate, you could add in the trading costs.

The Vanguard S&P500 index fund has a 6% turnover per year and pays 2 cents per share in brokerage fees.

If you assume that the average S&P500 stock has a $40 share price and a 1/2 point bid/asked spread, this would be the trading costs on a $100,000 investment.

6% x $100,000 = $6,000 worth of stock bought and sold

$6000/$40/share = 150 shares purchased

150 shares x $.50 = $75 in bid asked spread

brokerage commission = 150 shares x 2 cents = $3

$75 + $3 = $78 x 2 (to cover both buy and sell) = $166

Annual mutual fund mgt fee = $100,000 x 0.18% = $180

Total expenses = $166 + $180 = $346 -- 0.35%

As a rough rule of thumb, a well-managed index fund's trading costs are about equal to the annual expense ratio.

intercst

 

[malakito]

Re: Do you net out investment expenses from your S

If you want to get really accurate, you could add in the trading costs.

intercst

I'm still trying to figure out why someone would be tracking either of these costs.

I mean, I understand they're real and all that, but the only reason I can think that you'd do something like this is if some part of your withdrawal strategy depended on the performance of the mutual fund.

It seems to me that the way to do this would be to take the NAV at the end of the period and divide by the NAV at the beginning of the period.

Both the mutual fund operating expenses and trading costs are already accounted for at that point, though, right?

confusedly,

malakito.

 

[intercst]

Re: Do you net out investment expenses from your S

I'm still trying to figure out why someone would be tracking either of these costs.

I mean, I understand they're real and all that, but the only reason I can think that you'd do something like this is if some part of your withdrawal strategy depended on the performance of the mutual fund.

It seems to me that the way to do this would be to take the NAV at the end of the period and divide by the NAV at the beginning of the period.

Both the mutual fund operating expenses and trading costs are already accounted for at that point, though, right?

confusedly,

malakito.

THe SWR study is based on an S&P500 index that ignores investment expenses. We don't have 130 years of actual mutual fund data to base the study on, so we have to add in our best estimate of annual mgt expenses and trading costs to make the results more closely resemble reality.

Note: Trading costs for an actively managed mutual fund might be 3% or 4% per year if they buy and sell a lot.

intercst

 

[ESRBob]

Re: Do you net out investment expenses from your S

I'm still trying to figure out why someone would be tracking either of these costs.

Malakito,
Just to expand on Intercst's reply, since this is a key point: all the SWR analyses (pretty much) are assuming you are getting the full return of the asset class over time. You may have an index fund, but the fees mean you get that much less than the actual return of the asset class. That must then be subtracted from the SWR amount -- you need to pay your fund manager first.

If you are not using index funds, but have some sort of other diversified group of investments in an asset class, (say small stocks or US Large Growth), then you still need to assume the fees are subtracted from the returns historically due that asset class. (Anyway, that was all my original question, and it seems the board and Bernstein agree it is the correct methodology)

The really insidious thing is that in the real world, returns are up and down and all over the map, so you think you are getting the returns of the asset class over time, but you are getting the return less fees.

Intercst, thx for working out the math on the trading costs -- they can really zing you on an active fund! I guess the methodology of counting both buy and sell trade expense is accurate?

I'm humbled that you guys are getting your average fees down so low. I've been thinking I was doing ok to be at 40 basis points (which includes lots of exotic stuff like overseas, small, value, commodities, real estate etc) but still, that is a big enough difference to make me want to re-think this a bit.

But does anyone else share my aversion to the S&P500? I feel I got burned by it in 2000 when it had become so pumped up with just a handful of tech megacaps that I sort of swore off it (and at it!) and looked for large cap value index funds for the US Large Cap portion of the portfolio. I started thinking S&P500 could be thought of as a megacap growth fund. Any reactions?

thx
ESRBob

 

[malakito]

Re: Do you net out investment expenses from your S

Malakito,
Just to expand on Intercst's reply, since this is a key point:  all the SWR analyses (pretty much) are assuming you are getting the full return of the asset class over time.  You may have an index fund, but the fees mean you get that much less than the actual return of the asset class.  That must then be subtracted from the SWR amount -- you need to pay your fund manager first.

...

I'm humbled that you guys are getting your average fees down so low.  I've been thinking I was doing ok to be at 40 basis points (which includes lots of exotic stuff like overseas, small, value, commodities, real estate etc) but still, that is a big enough difference to make me want to re-think this a bit.

...

ESRBob

ESRBob and intercst,

Thanks for the replies. I had to think about it some more to figure out why I consider this a non-issue.

In my case, most of my money is in direct-ownership common stock accounts, which as a buy-and-hold my annual expense ratio is perhaps 1 basis point, or in very low cost index funds. My 401(k) is in a 5 basis point fund, I also own some VFINX, VTSMX, and the biggest expense ratio of them all, VPMCX. As mentioned above, my current expense ratio is around 15 basis points and dropping as I contribute new funds disproportionately to my lower-cost options. I also have to figure in some low balance fees on my kids' college funds as I switch from UTMA accounts to ESA accounts.

As an aside, to give an example of what I alluded to earlier, I just checked Vanguard's website and looked at the 10 year performance of VFINX versus the S&P500: 11.26% versus 11.34%, which is a haircut of 8 basis points whereas they listed their expense ratio as 18 basis points. That's as of 5/31/04, by the way.

In my own spreadsheets, I calculate what my withdrawal rate will be as follows:

1. I take my current nest egg and add in projected future contributions and projected returns.
2. I take my current expenses (monthly average over the last six months) and add in projected inflation.
3. I take #2 and divide it by #1.

I have this projected on a monthly basis from now until 2025, and figure that when the number starts getting close to 4%, I'll look more closely at it.

At any rate, I believe that out in that 202X timeframe, my withdrawal rate will be dropping by around 9 basis points per month. In other words, this discussion only affects my retirement date by one month. One month out of 20 years is not really a big deal. IOW, IMHO, this is down in the noise.

I do agree that for someone whose investment costs run 3-4% annually, it's a much bigger deal. Also, it makes me realize that I need to do something about my VPMCX holdings -- either stop adding to them or convince my wife to move them to a lower cost fund.

malakito.

 

[cute fuzzy bunny]

Re: Do you net out investment expenses from your S

But does anyone else share my aversion to the S&P500? I feel I got burned by it in 2000 when it had become so pumped up with just a handful of tech megacaps that I sort of swore off it (and at it!) and looked for large cap value index funds for the US Large Cap portion of the portfolio. I started thinking S&P500 could be thought of as a megacap growth fund. Any reactions?

The s&p500 scares the crap out of me at current valuations. I'm buying nothing but value stocks at PE's in the 14 range, and not a lot of those either.

 

[BigMoneyJim]

Re: Do you net out investment expenses from your S

But does anyone else share my aversion to the S&P500? I feel I got burned by it in 2000 when it had become so pumped up with just a handful of tech megacaps that I sort of swore off it (and at it!) and looked for large cap value index funds for the US Large Cap portion of the portfolio. I started thinking S&P500 could be thought of as a megacap growth fund. Any reactions?

I must not be seeing things the same way, but I don't claim to be a sophisticated investor. Chickbull has pooh-poohed the TSM or S&P500 a couple of times recently for it's bad 2000-2003 performance, and I don't get that, either.

My question is: what are the alternatives? Do you want to avoid growth stocks altogether? Did you want to be able to rebalance that sector? How much better off do you think you would have been? What about the extra trading costs and/or higher fees?

I'm convinced that if I actively managed my investment allocations based on my perception of valuation I would perform far worse than TSM or S&P 500. Yeah it hurt to see my portfolio go down by 1/3 or more, but I'm sticking through it. (I am accumulating, though--not withdrawing.) Furthermore, I still haven't lost principle; I just lost some of my gains from the previous 3 or 4 years, but I still have way more than I put in.

Maybe it's just me, but almost every argument against TSM or S&P 500 seems to boil down to market timing and/or judging valuation. Quite a few people were right in 1999-2000 about the tech market being way overvalued; if you were one of them, congratulations, but I have to ask this: How many other times were you wrong, and overall would you have come out ahead shifting allocations based on your market judgement?

 

[sgeeeee]

Re: Do you net out investment expenses from your S

. . .
I'm convinced that if I actively managed my investment allocations based on my perception of valuation I would perform far worse than TSM or S&P 500.
. . . Maybe it's just me, but almost every argument against TSM or S&P 500 seems to boil down to market timing and/or judging valuation.

Hi BMJ,

You sum up my feelings pretty well too. I wish I had something better (more return with less risk) but it just doesn't work that way.

 

[cute fuzzy bunny]

Re: Do you net out investment expenses from your S

Dear Debt Free Jim...

Well, it does come down to valuations/timing. Some people just cant help themselves.

With regards to rightness, by being right and getting out in 1/2000 I enabled an ER. As I've mentioned though, I wanted to get out in late 99 and only hung in there to defer the capital gains taxes for another year. Had I gotten out when I wanted to, at nasdaq ~ 4000, I would have missed another 1000 points of insane upside.

As far as rightness the rest of the time..well, mixed results. I seem to be good at judging big indexes, the bigger the better. I seem to be bad at judging sectors and individual stocks. So while I have a nice 7 figure portfolio to point at proudly, I also have a nice 6 figure capital loss I'm dragging behind me. Which does help me not pay any taxes and will continue to do so for a good number of years.

Some value stocks do offer a good dividend and potential for some upside with limited downside. Its hard to beat up a well known sturdy company paying a nice dividend that only has a PE of 10 or 12 or 14. Thats where I am for now.

 

[ESRBob]

Re: Do you net out investment expenses from your S

BMJ and TH,
Actually my point was slightly different, but I think in an important way.

I wasn't making any claim as to whether the SP500 was overvalued or undervalued itself, now. I am objecting to the index itself as giving a false sense of diversification -- you think you are buying 500 stocks, but in reality, (I will check the figs for a followup post) since it is a cap-weighted index, you are getting a huge proportion of your SP500 investment in a few dozen companies, and almost nothing in the bottom300 or 400 stocks.

That is my objection, that while it may be an accurate capweighting of the stock market, it is not really giving you diversifcation into small or even medium-sized stodks and isn't ever doing a very good job diversifying you around the large caps.

Then, BMJ, you 'outed' me on another bias, though, which is growth vs value. I've been reading enough stuff by the DFA and Chicago crowd, (Fama/French) to make me think that at a minimum a value 'tilt' is a big advantage in virtually any stock portfolio (they reckon some sort of persisting value premium of 2-3% per year in small, large, international, domestic --anything)

Growth stocks just seem to make lousy investments. Near as I can explain it to myself, if everybody loves a company, then its stock is bid up. If everybody knows a company sucks, then its stock is beat up. You may want to buy the growth company's products, but the slightest hiccup will trash its stock. Avoid the value company's products perhaps, but if you have a diversified batch of their stocks (obviously any single stock can go bankrupt, they are bid down for a reason!) then the stocks will eventually come around, and statistically much better than growth stocks in performance. Just not falling off a cliff means their prices will tend to get bid up when people forget they are basket cases.

The SP500 tends to have a growth stock bias, too, if only because only about 20% of stocks might be reasonably considered value stocks. You can still get a value index, value doesn't have to be actively managed.

ESRBob

 

[cute fuzzy bunny]

Re: Do you net out investment expenses from your S

Ah yes. They should call it the S&P10+490.

The value vs growth premium is simple: people will pay a big risk premium for troubled companies, and few really drop dead and a lot actually come back. On top of that they usually pay a nice dividend.

The growth thing is explicable: nobody stays on top of the heap for long. Who are the power players in a particular market today? Who were they yesterday? Day before?

Look at tech: intel, microsoft, dell today. HP, Digital, Wang yesterday. IBM, burroughs, sperry the day before.

Its just too hard to stay on the heap, and nobody wants to pay a premium for those on top of it!

 

[malakito]

Re: Do you net out investment expenses from your S

BMJ and TH,
I wasn't making any claim as to whether the SP500 was overvalued or undervalued itself, now.  I am objecting to the index itself as giving a false sense of diversification -- you think you are buying 500 stocks, but in reality, (I will check the figs for a followup post) since it is a cap-weighted index, you are getting a huge proportion of your SP500 investment in a few dozen companies, and almost nothing in the bottom300 or 400 stocks.
ESRBob

According to http://www.decisionpoint.com/ChartSpotliteFiles/030917unweighted.html, "For example, about 70% of the S&P 500 weighting is in the 50 (10%) largest stocks, so claims by many financial advisors that investing in an S&P 500 Index fund diversifies your portfolio over 500 stocks is simply not true."

A couple of points:

1. Cap weighting in an S&P 500 index fund reduces trading costs dramatically.

2. Historical data for non-capitalization weighted indices are, as far as I know, non-existent. Unless you want to go with the DJIA, which is even fewer stocks (30 currently, as few as 12 in the past I believe).

3. 50 stocks is still a lot of diversification.

4. Any other large stock mutual fund will end up cap-weighted to some degree simply because they are limited to only buying a certain percentage of any given company. With as much money as these large funds have, they can't really play in the small market-cap arena -- they'd end up having to buy the entire company.

malakito.

 

[cute fuzzy bunny]

Re: Do you net out investment expenses from your S

I have seen some non-cap weighted data at some point, dont remember where, but the non cap weighted did outperform cap weighted.

The biggest problem you'd have is if all the s&p500 funds werent cap weighted, the bottom 50-70 companies would have to multiply their stock issuance by an order of magnitude to have enough shares just for the index funds!

 

[ESRBob]

Re: Do you net out investment expenses from your S

Here's a link to the cumulative weightings on the S&P 500.
http://www.indexarb.com/indexComponentWtsSP500.html

As long as you know what you're getting, its fine. I want to be widely diversified, and for me that has meant stepping outside of simply matching the cap weighted market. I don't want to tie my fortunes so closely to 50 firms or worse, the top 15 firms which make up 30% of the index.

After all, we are seeking less-correlated or non-correlated asset classes, and the SP500is inherently yahnked around by the fortunes of those 15 or 50 companies, which often tend to move in lockstep with the fortunes of the overall economy. I do hold a bit of the SP 500 index for convenience, but my large cap stuff is loaded up with Value Index (and yes, I admit it, I have some of Vanguard's Primecap leftover from my earlier unenlightened pre-index days). But for stocks overall, we are very heavily loaded up with small cap (again with a value tilt)-- they have good long term performance and are low-correlation to the big stocks and just about everything else. (albeit with high volatility).

I am struggling to try to express this concept, and I know that lots of smart people are lined up against me on this one -- S&P 500 is the supposed god of prudent diversification, but I am just feeling strongly like there is a better way to diversify, perhaps at marginally higher cost, but with much better long run performance. Can anybody help me out here? I'm not even sure how to do a study to shed light on this, but I'm thinking it is important.

 

[charlie]

Re: Do you net out investment expenses from your S

ESRBob,

I compared the S&P500 Index to the equity part
of the "coffeehouse" portfolio, which is composed
of equal weightings of S&P500, Large Cap Value,
Small Cap Value, Small Cap Index, Total International
and REIT Index for the 20 year period from 1984
through 2003. The REIT Index was only included
from 1997 through 2003. The results were:

S&P500: Total annualized return = 12.78%, with
sigma = 16.65%

Coffeehouse : Total annualized return = 12.26% with
sigma = 14.61%

The most interesting thing I discovered was that
the Coffeehouse had returns of 4.50%, -2.48%,
and -14.77% vs. S&P500 with -9.06%, -12.02% and
-22.15% in 2000, 2001 and 2002 respectively.
The REIT Index, Small Cap Index and Small Cap
Value really helped in that time period.

What I took away from this is that diversification
truly helps dampen volatility even if long term
returns for each asset class are very similar.

Cheers,

Charlie

 

[BigMoneyJim]

Re: Do you net out investment expenses from your S

I wasn't making any claim as to whether the SP500 was overvalued or undervalued itself, now. I am objecting to the index itself as giving a false sense of diversification -- you think you are buying 500 stocks, but in reality, (I will check the figs for a followup post) since it is a cap-weighted index, you are getting a huge proportion of your SP500 investment in a few dozen companies, and almost nothing in the bottom300 or 400 stocks.

Okay, that's a good point, and one I had forgotten. I'll roll that around in my head for a while and see what comes out.

Then, BMJ, you 'outed' me on another bias, though, which is growth vs value. I've been reading enough stuff by the DFA and Chicago crowd, (Fama/French) to make me think that at a minimum a value 'tilt' is a big advantage in virtually any stock portfolio (they reckon some sort of persisting value premium of 2-3% per year in small, large, international, domestic --anything)

Growth stocks just seem to make lousy investments. Near as I can explain it to myself, if everybody loves a company, then its stock is bid up. If everybody knows a company sucks, then its stock is beat up. You may want to buy the growth company's products, but the slightest hiccup will trash its stock. Avoid the value company's products perhaps, but if you have a diversified batch of their stocks (obviously any single stock can go bankrupt, they are bid down for a reason!) then the stocks will eventually come around, and statistically much better than growth stocks in performance. Just not falling off a cliff means their prices will tend to get bid up when people forget they are basket cases.

My intuition--or perhaps my 2000 experience--has me wary of growth stocks, too. I don't have a strong opinion--just a feeling. . . .

Sorry, this post got interrupted by some phone calls a couple of hours ago and I have long since lost my train of thought. I will give some more thought about the weighting of S&P 500 and TSM (I'm assuming your objections to S&P 500 apply almost as well to TSM) growth stocks versus value stocks and plausible (for me) alternatives.

 

[wabmester]

Re: Do you net out investment expenses from your S

S&P 500 is the supposed god of prudent diversification, but I am just feeling strongly like there is a better way to diversify, perhaps at marginally higher cost, but with much better long run performance.

Who made the S&P 500 god? It's more like the model-T of indices. If you want true diversity, own the entire market. If you believe the market is efficient, then own the entire market using cap weighting. I think the best you can do from a diversity and cost standpoint is 50% Vanguard Total Stock Market + 50% Vanguard Total International Stock Market.

Adding a "value tilt" is similar to buying a junk bond fund. You get a risk premium, but you also get high volatility. You mitigate risk by buying lots of junk and hoping that some of it becomes quality. A perfectly reasonable strategy.

 

[cute fuzzy bunny]

Re: Do you net out investment expenses from your S

OOOh Wabmester, I disagree.

Owning a broad collection of value stocks, either hand picked, in a manged fund specializing in value stocks, or in a broad index does NOT appear to me to induce more volatility as a rule.

In fact, in times of overvaluation like these, there is some comfort in owning a collection of stocks with price to earnings ratios in the single and low double digit range.

To be fair, owning a few value stocks can create tremendous volatility. If you have a fund or a collection of 50 or more well chosen issues, I think you have little to worry about.

However, I've yet to see any analysis of value stocks that measured longer then 3 years that didnt put value well ahead of growth or blends. Heck, Bernstein devoted an entire section of the four pillars to the curious aspect of crappy companies producing consistently superior results...