Motif Investing - own the S&P500 for less

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
conlius
Posts: 92
Joined: Fri May 02, 2014 11:09 am

Re: Motif Investing - own the S&P500 for less

Post by conlius » Wed Aug 27, 2014 3:53 pm

Problem is, it's not good for a regular investor. The commission is the primary problem. Someone buying every month is going to lose a lot in commission over time. Commission is not a concern when buying from many other brokers.

The primary thing I thought was cool about Motifs when i read into them was that you could invest in an idea or specific strategy. Invest in solar power companies, caffeine companies, invest in companies that invest in private equity, or invest like . I'm still considering a small portion, say 3-4% of my portfolio as the "fun" part. I also have another 3-4% allocated purely to option strategies. My core, which makes up 90%+ of my portfolio, will stay with Vanguard.

User avatar
JoMoney
Posts: 6081
Joined: Tue Jul 23, 2013 5:31 am

Re: Motif Investing - own the S&P500 for less

Post by JoMoney » Wed Aug 27, 2014 4:04 pm

inbox788 wrote:...I was wondering what the outcome could be if you randomly selected 30 stocks in the sp500 (sampling; either market weighed or equal weight) and invested in them for 10 years or longer (1 motif). Expense wise, it could be lower than an index fund. Performance wise, I think things will average out over the years. Now if I did this with another 30 stocks each year (maybe choosing from all 500 again, or excluding those that are already in the portfolio), I'd be getting closer to the sp500 each year I did this. Measured over decades, will the cost savings outweigh the skewness? After 20 years x 30 stocks a year = 600 stocks, I'd have opportunity to do a little rebalancing if I wanted to be active. Or just roll the dice passively. Will have to come up with a simple algorithm to weigh and adjust the contributions so the early contributions (year 1) don't over shadow the later ones (year 30). Maybe buy a motif every $10k? Adjust for market growth?
You may be interested in looking at the performance of the LEXCX fund over the years. In the 1930's they took an equal weighting of 30 of the top dividend paying stocks on the NYSE and have never traded the portfolio since. Just "buy and hold" and no rebalancing of the consituents, buying/selling of new flows in/out the fund just goes to the exact weighting as it exists. The 21 constituents today are the result of various mergers and split-offs etc.. It's had long periods of under-performance, more recently outperforming the market, overall very much market like performance. I like to compare it to the Fidelity fund FFIDX which has had performance very similar to the S&P500 and has a history going back to the 1930's as well.
Morningstar Chart Link
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham

mcblum
Posts: 464
Joined: Wed Jan 06, 2010 7:46 am
Location: gaithersburg,md

Re: Motif Investing - own the S&P500 for less

Post by mcblum » Wed Aug 27, 2014 4:10 pm

I am to lazy for this. To much work!
Marty

Wagnerjb
Posts: 7203
Joined: Mon Feb 19, 2007 8:44 pm
Location: Houston, Texas

Re: Motif Investing - own the S&P500 for less

Post by Wagnerjb » Wed Aug 27, 2014 4:31 pm

Tanelorn wrote:I agree that at the 5bp level, you'll probably not get many takers for the time-vs-reward tradeoff.
Agreed. In my opinion, the only way this makes sense is for the tax benefits.
With the updated TLH value being worth ~50bp extra (not included in my original estimates), now you're talking $5k/year (plus the other $500 in ER savings), so that's potentially worth the time and hassle if it's not too much trouble.
Yes, but don't overlook the other tax benefits that an individual stock portfolio can offer. You can cherry pick the biggest winners and donate them to charity - for the most tax efficient charitable contributions. You can cherry pick the biggest winners and gift them to your children - and pay 0% tax on those gains. You can cherry pick the losers and TLH them. Those stocks in the middle are perfect for those who retire early. You have a period between retirement and social security which is prime territory for converting IRAs to a Roth in a low tax bracket. To keep your tax bracket low, you will sell some of these individual stocks for spending money....those with smaller gains.

Another potential benefit is the lack of mandated turnover in your portfolio. If you buy all 500 stocks, you are forced to follow this turnover. But if you buy a diversified sample of large cap stocks you have lower turnover than the index. AND...you avoid the 25bps of reduced returns due to front-running of the S&P500. That means your costs are lower (less turnover) and your returns are higher (no frontrunning).

Between the benefits of TLH, the absence of an ER, the absence of front-running, and the more tax efficient donations and gifting....you are getting into total benefits that are substantial. Are they worth your time and effort to achieve these? Only if you have the knowledge, the portfolio size and the patience to do this.

Best wishes.
Andy

placeholder
Posts: 3958
Joined: Tue Aug 06, 2013 12:43 pm

Re: Motif Investing - own the S&P500 for less

Post by placeholder » Wed Aug 27, 2014 4:43 pm

Tanelorn wrote:So yes, you might have to pay to sell them all eventually, but that's a long way down the road and brokerage commissions have been dropping steadily. There are brokers offering lots of free trades, both per year, and for new customers, at the $100k asset level where this starts to make sense so I think it's not unreasonable to assume you wouldn't take a big loss on this.
Even easier at TD Ameritrade:
New accounts funded with $2,000 or more are eligible for 500 commission-free trades for 60 days.

User avatar
ogd
Posts: 4854
Joined: Thu Jun 14, 2012 11:43 pm

Re: Motif Investing - own the S&P500 for less

Post by ogd » Wed Aug 27, 2014 8:45 pm

Tanelorn wrote:Thank you bpp for the link. The TLH benefit is cited from an academic paper simulating holding 500 stocks instead of a 500 fund at around 0.5% annually compared to doing nothing. Obviously a fund holder would get some benefit from TLH, but only during market crashes. I would guess the vast majority of the 0.5% TLH benefit would apply in this case relative to a fund/ETF holder. The study assumed a 35% marginal tax rate, so it could be scaled up or down by 50% per individual circumstances.

http://www.firstquadrant.com/downloads/ ... t_Loss.pdf

This makes the annual benefits about 10x higher than the 0.05% cost savings I estimated, or $500/year/$100k invested. This seems meaningful.
Not so fast. There are problems with this number, as applied to motif investing. One is simple: 30-stock motifs are far less volatile than individual stocks, even if you group them by sectors or whatever. So there just won't be as many harvest opportunities. Another is:
Firstquadrant wrote:Each year, we take any tax obligations out of each portfolio and any tax savings from loss harvesting are reinvested back into the portfolio.
, which you can't do without messing up your costs. This is important, because it (and presumably dividends, which they don't talk about directly) creates high tax lots on an ongoing basis; if you don't have those, harvesting tapers off. In fact, the very act of harvesting at costs significantly higher than zero might take away most of the alpha.

Another problem is why I would not do loss harvesting with individual stocks, like I mentioned, even if I otherwise subscribed to creating-my-own-index:
Firstquadrant wrote:We ignore the wash rule for purposes of the simulation(2). The wash rule typically has a slightly negative impact in the real world because of the mild tendency for sharply-fallen assets to rebound slightly, but this will not significantly change our results.

(2)The wash rule states that assets cannot be repurchased within one month of their sale in order to receive the tax credits. It matters a great deal in the real world. In a monte carlo simulation, in which each asset has a random return, with the same distribution as any other asset, it does not make any difference.
I'm sorry, but I refuse to sell a group of bad performing stocks and buy average stocks instead. As far as I know, this is one of the few ways to consistently lose money in the market, see behavioral gap. Motifs don't help you here either, because tax lots / wash sales happen at the level of individual stocks, so you can't use the old trick of using a heavily overlapping but not identical fund.

bpp
Posts: 1964
Joined: Mon Feb 26, 2007 12:35 pm
Location: Japan

Re: Motif Investing - own the S&P500 for less

Post by bpp » Wed Aug 27, 2014 9:48 pm

ftobin wrote:
bpp wrote:Actually, the real secret is that there is no point in slavishly trying to track any index. The fact is that if one has enough separate holdings, it really doesn't matter too much what exactly those holdings are, or in what weights. One will end up tracking the overall market regardless. (Modulo, say, a small-cap tilt if selecting stocks randomly without regard for capitalization.) So spending time worrying about weights and tracking error is pretty much unnecessary.
I think following an index would be important, else how do you measure whether or not you are being successful at being cheaper than paying Vanguard a small amount to handle it for you? If you aren't following a measuring stick, any divergence could be attributed to exposure to different risk factors.
In my case, I'm not trying to undercut Vanguard. (They don't even offer a Japan-only index.) I'm trying to avoid dealing with PFIC issues, and that has been very successful for the past 10 years. Not getting whacked with mark-to-market taxes on last year's rocketship-ride in Japanese stocks alone has probably more than paid for the strategy. The TLH harvesting is a bonus, and one that I've only recently really started appreciating the value of.

That said, I have tracked my portfolio against Japanese indices: Large-caps against the Nikkei, small-caps against the Tokyo Stock Exchange Section 2 Index, and the overall portfolio against the TOPIX. I find that the large and small cap portions track their respective benchmarks well enough, and divergences from the TOPIX index are largely due to whether large- or small-caps did better in a given year. (I divide my stocks 50/50 large and small caps). I'm not trying to measure effects at the few-bp level, though. Just trying to make sure that I'm not getting wildly clobbered by skewness effects, which I'm satisfied by now is not the case.

If one did want to measure tracking error at the few-bp level, it would take a lot of work, accounting precisely for dividends, etc. And trying to analyze how much tracking error could be attributed to what factor exposures would be an even bigger job. Not clear to me that it would be all that productive an endeavor, though. In the OP's case, the main benefits would come from TLH, not expense-ratio reduction, and the benefits of TLH can be measured directly each year. If one trusts the markets to be relatively efficient, that seems enough for most practical purposes to me.

inbox788
Posts: 5635
Joined: Thu Mar 15, 2012 5:24 pm

Re: Motif Investing - own the S&P500 for less

Post by inbox788 » Wed Aug 27, 2014 11:00 pm

ogd wrote:Not so fast. There are problems with this number, as applied to motif investing. One is simple: 30-stock motifs are far less volatile than individual stocks, even if you group them by sectors or whatever. So there just won't be as many harvest opportunities.
...I'm sorry, but I refuse to sell a group of bad performing stocks and buy average stocks instead. As far as I know, this is one of the few ways to consistently lose money in the market, see behavioral gap. Motifs don't help you here either, because tax lots / wash sales happen at the level of individual stocks, so you can't use the old trick of using a heavily overlapping but not identical fund.
I think I've figured out a passive scheme that might work to reduce investment fees for the accumulation phase, say a young investor that approaches the performance of an equal weight sp500 fund (no rebalancing). Start with the first $10k invested in 30 random sp500 stocks as motif1. You'd start with around $333 in each of the first 30 stocks. By the time you've ready to invest the next $10k, the market might have risen 10%, so you need to wait till you have $11k to invest and choose a second motif2 with 30 other random stocks. Your first motif1 should more or less have risen about 10% during this time. By the time you're ready to invest your next $10k or so, sp500 might be up 20%, so again, wait till you have $12k to invest in the third lot motif3. It's as if you've bought all these stocks at around $333 in original year 1 values, adjusted for average market gains (losses) over the years. By the time you average motif1, motif2, motif3, etc. the volatility and variance between your motifs and the equal weight sp500 should diminish. As you include more sp500 stocks over the years/decades, going forward, you'll pretty much build a fund with an expense of 0.1% or less over the holding period of the motif, which may be years/decades, starting with $10/10,000+ and decreasing as the market goes up.

As you use the funds, you could do FIFO on the motifs, or try to minimize taxes, or if there are any lots with losses, try to harvest losses, but probably unlikely over longer periods. If the motifs have high variability, and you sell losing lots, you may be selling bad performing stocks and keeping average ones. But given number of stocks in each motif and long periods, I expect most motifs will cluster around the sp500 average, and a few do a little better and a few do a little worse. I think you can save about $50 in fees for every $10k invested/year, so after a decade or two that could be $500-1000/motif, so after 20 years, an extra $5000 in fees might be saved.

TLH will burn fees, so they're work opposite of saving fees. If you plan to cash out, and eventually pay taxes, I don't see tax deferment as a huge savings, unless you have lower tax rate in the future years. Donating and getting step up basis is a valuable benefit, but that means someone else benefits.

If wash sales are passed through, it would be a nightmare if you invested in others motifs,and you had to track all the overlapping stocks that could lead to a wash sale. A motif could make money, but one underlying stock loses money and happens to be in the new motif you just bought. Wash sale!

User avatar
ogd
Posts: 4854
Joined: Thu Jun 14, 2012 11:43 pm

Re: Motif Investing - own the S&P500 for less

Post by ogd » Wed Aug 27, 2014 11:39 pm

inbox788 wrote: I think you can save about $50 in fees for every $10k invested/year, so after a decade or two that could be $500-1000/motif, so after 20 years, an extra $5000 in fees might be saved.
Well almost, except it's $5 per $10k per year, which in recent years seems to have shrunk down to nearly nothing (whether it's the security lending, better trading avoiding S&P front running, or something else, I don't know). Does that still sound enticing?
inbox788 wrote:If wash sales are passed through, it would be a nightmare if you invested in others motifs,and you had to track all the overlapping stocks that could lead to a wash sale. A motif could make money, but one underlying stock loses money and happens to be in the new motif you just bought. Wash sale!
Wash sales / tax lots are passed through, because there is no underlying security. So it is a nightmare if you ever have to dive down to individual lots as opposed to aggregates.

Look, I'm not trying to put down the whole idea of motif investing. It could be great for stock pickers / people with theses they want expressed / people in unusual investment circumstances for whom no good fund exists. But when an insanely cheap broad fund already exists that simplifies things to such a degree, why would you promote motifs as an alternative, which I don't think they were designed to do?

The other thing is, for larger portfolios you can get really low brokerage fees anyway, e.g. I get 25 free trades / quarter and $2 after that, and I'm told that at Wells you can get all free trades or something of the sort. So I could juggle individual stocks if I really wanted to avoid the fees, but it's still not worth it.

Tanelorn
Posts: 1546
Joined: Thu May 01, 2014 9:35 pm

Re: Motif Investing - own the S&P500 for less

Post by Tanelorn » Thu Aug 28, 2014 2:23 am

ogd wrote:
Tanelorn wrote:This makes the annual benefits about 10x higher than the 0.05% cost savings I estimated, or $500/year/$100k invested. This seems meaningful.
Not so fast. There are problems with this number, as applied to motif investing. One is simple: 30-stock motifs are far less volatile than individual stocks, even if you group them by sectors or whatever. So there just won't be as many harvest opportunities.
I agree that their assumptions should be checked and one should decide how much this benefit is likely to be in practice vs in theory. However, your concern about less volatile baskets isn't one of them - you harvest losses at the individual stock level. For $5 you can sell a security out of your Motif basket and then the same to buy it back.
Firstquadrant wrote:Each year, we take any tax obligations out of each portfolio and any tax savings from loss harvesting are reinvested back into the portfolio.
, which you can't do without messing up your costs. This is important, because it (and presumably dividends, which they don't talk about directly) creates high tax lots on an ongoing basis; if you don't have those, harvesting tapers off. In fact, the very act of harvesting at costs significantly higher than zero might take away most of the alpha.
It doesn't seem to me like dividends or tax savings reinvestment is the main source of TLH, I thought it would be the skewness of returns over the long term, but I could be wrong. In practice reinvesting tax savings is easy (reduce your withholdings and you've got more cash to invest), but it does put these savings into the "small lump sum" bucket like dividend reinvestment that will not be cost effective to buy into Motifs. Basically this means that the TLH savings, say 0.5%/year or whatever, aren't going to compound with the same gains since they'll get dumped back into your ETF or fund holding until you've got a big enough lump sum to switch to Motifs again.
I'm sorry, but I refuse to sell a group of bad performing stocks and buy average stocks instead. As far as I know, this is one of the few ways to consistently lose money in the market, see behavioral gap. Motifs don't help you here either, because tax lots / wash sales happen at the level of individual stocks, so you can't use the old trick of using a heavily overlapping but not identical fund.
I'm not sure this "investor returns vs fund returns" difference can be properly attributed to the issue of the rebounding of fallen individual stocks. First, I think the whole fund-vs-investor returns is often a time-weighted vs dollar-weighted return issue, where as funds grow much bigger and more popular based on their early good performance, they get bloated and perform less well. This means the average dollar came in much later and earned the later, poorer returns. It's true some of the effect may be bad investor behavior (trading the fund shares at a loss), but I don't think it's most of it.

Second, on the subject of rebounding at the individual stock level, that was so obvious, people would just buy any big drop and the market would be more efficient and stop behaving that way. Isn't this the opposite of the momentum factor says typically happens, that things that drop tend to keep on dropping for a while? Supposedly momentum is pretty robust in explaining returns, and if so, it would make sense to cut your losses and get a tax benefit rather than holding something that's likely to keep falling. Hopefully by the time of the negative momentum abates, the wash sale window will have past and you can buy it back without the negative short term expected return.

That said, what I do think can happen is that the market at large can sell off, people can panic and sell, and you can do a little better by buying the market during these dips. However, if what you mean by a rebound effect applies only the aggregate market level rather than the individual stock level, we won't miss out on average if we take our individual TLH losses and reinvest them in VOO in the meanwhile.

Wagnerjb
Posts: 7203
Joined: Mon Feb 19, 2007 8:44 pm
Location: Houston, Texas

Re: Motif Investing - own the S&P500 for less

Post by Wagnerjb » Thu Aug 28, 2014 8:09 am

inbox788 wrote:TLH will burn fees, so they're work opposite of saving fees.
TLH is free if you don't pay commissions...and many people don't pay commissions these days. As a Flagship member at Vanguard, I sure don't.
If you plan to cash out, and eventually pay taxes, I don't see tax deferment as a huge savings, unless you have lower tax rate in the future years.
So, you must not want an IRA or a 401k plan if you don't feel that tax deferment is a big deal. But even if you don't appreciate the benefits of tax deferment, it isn't a stretch to think that many people will be in lower tax rates in their retirement years. I will be.
Donating and getting step up basis is a valuable benefit, but that means someone else benefits.
No. Donating means that I NEVER pay those taxes, while the guy with the mutual fund does. If I own two stocks - one with a $5000 gain and one with a $5000 loss - I can harvest the loser and save real money in taxes. Then I can donate the winner and not pay tax on that money. The guy who owns those two stocks in his mutual fund cannot do either. He doesn't get the cash benefits of TLH and he has to work harder to generate enough to donate $5000 to his favorite charity.

Best wishes.
Andy

HenryPorter
Posts: 161
Joined: Sat Dec 05, 2009 2:09 am

Re: Motif Investing - own the S&P500 for less

Post by HenryPorter » Thu Aug 28, 2014 9:34 am

----------------------------
Last edited by HenryPorter on Sun Sep 14, 2014 10:02 pm, edited 1 time in total.

User avatar
ogd
Posts: 4854
Joined: Thu Jun 14, 2012 11:43 pm

Re: Motif Investing - own the S&P500 for less

Post by ogd » Thu Aug 28, 2014 12:48 pm

Tanelorn wrote:I'm not sure this "investor returns vs fund returns" difference can be properly attributed to the issue of the rebounding of fallen individual stocks. First, I think the whole fund-vs-investor returns is often a time-weighted vs dollar-weighted return issue, where as funds grow much bigger and more popular based on their early good performance, they get bloated and perform less well. This means the average dollar came in much later and earned the later, poorer returns. It's true some of the effect may be bad investor behavior (trading the fund shares at a loss), but I don't think it's most of it.
The time-weighted vs dollar-weighted difference only comes into play when looking at the results of a single fund. But when doing aggregate numbers, you are comparing two dollar-weighted figures. Everyone will tell you that this is a real effect of investors jumping in the wrong direction at the wrong time and not a mathematical effect like "contributing to the wrong average" or something of the sort.

It's true that some of the gap is because of investors jumping asset classes entirely, e.g. selling out of stocks in a crash, which is not applicable to our problem. I don't think it's all, though, and I do think that mean reversion exists; the Firstquadrant paper acknowledges it too as a negative. As for momentum, I've never been able to reconcile it with mean reversion, or efficient markets, or anything else for that matter. For my part, I choose to believe that selling depreciated assets without replacing them with highly correlated ones (more correlated than possible with single stocks) is a bad idea and my TLH, like many others, is an exercise in juggling correlated funds. We're also getting rather off-topic; suffice it to say that I think that a 0.5%/year advantage from harvesting is huge, needs much more substantiation and probably its own discussion focusing on whether to use individual stocks instead of mutual funds always.

inbox788
Posts: 5635
Joined: Thu Mar 15, 2012 5:24 pm

Re: Motif Investing - own the S&P500 for less

Post by inbox788 » Thu Aug 28, 2014 1:23 pm

ogd wrote:
inbox788 wrote: I think you can save about $50 in fees for every $10k invested/year, so after a decade or two that could be $500-1000/motif, so after 20 years, an extra $5000 in fees might be saved.
Well almost, except it's $5 per $10k per year, which in recent years seems to have shrunk down to nearly nothing (whether it's the security lending, better trading avoiding S&P front running, or something else, I don't know). Does that still sound enticing?
Oops, off by a decimal point. No, it was never all that enticing given all the risks and drawbacks, and now that it's an order of magnitude less, even less so. What is fascinating is that lower fees are possible than the lowest fees currently being charged by the likes of Vanguard, Schwab, and Fidelity. Much of the Motif cost savings comes from the flat fee with no annual charge, so you can view it as an illiquidity premium, small as it may be.

Austintatious
Posts: 773
Joined: Thu Sep 13, 2012 7:01 pm

Re: Motif Investing - own the S&P500 for less

Post by Austintatious » Thu Aug 28, 2014 1:35 pm

a 2012 Motley Fool article re Motif

http://www.fool.com/investing/brokerage ... ution.aspx

Buysider
Posts: 721
Joined: Sun May 18, 2008 9:36 pm

Re: Motif Investing - own the S&P500 for less

Post by Buysider » Thu Aug 28, 2014 1:50 pm

Much of the Motif cost savings comes from the flat fee with no annual charge, so you can view it as an illiquidity premium, small as it may be.
Investors focus on explicit costs because, well, they are explicit. Implicit, or hidden costs, are harder to compare and therefore are less easily analyzed. I think this thread is a little short on that.
Plus, they've got Goldman as their lead investor (article), which suggests the smart guys think they've got a good business. I think I know why too, but that's for another thread.
Do I think their firm has come up with a way to deliver trading more cheaply than etrade or fidelity, both of which have incentives to keep their prices below cost to aggregate assets and cross sell products? I don't - I think the company is making money selling equity to JPM and Goldman Sachs, with the hope that they gain scale and hype enough to take this public before the crashing reality that they can't make any money at $0.33 a share unless they really jack up how much they are scalping on the execution of the trades (the difference between good and bad execution, even on market orders, can be in the 50 bps range, which would negate this whole discussion thread).

In any event, the firm has already "pivoted" to try and be a bit more like wealthfront/betterment. Charging 15-40 bps for asset allocation and custody is a lot better than $0.33 per trade. Will be interesting to see how it ends.

berntson
Posts: 1366
Joined: Mon Oct 29, 2012 12:10 pm

Re: Motif Investing - own the S&P500 for less

Post by berntson » Thu Aug 28, 2014 2:04 pm

inbox788 wrote: I think I've figured out a passive scheme that might work to reduce investment fees for the accumulation phase, say a young investor that approaches the performance of an equal weight sp500 fund (no rebalancing). Start with the first $10k invested in 30 random sp500 stocks as motif1. You'd start with around $333 in each of the first 30 stocks. By the time you've ready to invest the next $10k, the market might have risen 10%, so you need to wait till you have $11k to invest and choose a second motif2 with 30 other random stocks...
Here is a similar strategy I've thought a bit about. Say that Alice is a new investor who plans to invest about $5,000 every month starting in January. She wants to hold individual stocks that more or less track the performance of the S&P. Her strategy is to buy companies with the largest market cap first. She starts by saving for two months and then buys $10,000 in Apple stock. In the fourth month, Exxon's market cap is 74% of Apple's market cap, so Alice buys $7,400 worth of Exxon stock, enough for it to be worth 74% as much as her Apple holdings. (This leaves a little cash left over, but that's alright. She can invest it next month.) On the fifth month, Alice buys $5,600 worth of Microsoft stock. On the sixth month, she buys and $4,900 worth of Johnson and Johnson stock. And so on and so forth until she has bought all 500 stocks in the index.

The beauty of this system is that there is never any need to rebalance. Over the course of ten to fifteen years (depending on market returns), Alice will build a 500 stock portfolio that exactly matches the S&P 500. And she can do it with exactly 500 purchases while investing new money. Interactive Brokers charges $1 commissions on stocks, so she will be paying $40-$50 a year in brokerage fees. Once Alice completes the process, she can start again by making a new purchase of Apple stock, then bringing all the rest of the ratios back into line.

asha1001
Posts: 39
Joined: Thu Mar 06, 2014 10:43 am

Re: Motif Investing - own the S&P500 for less

Post by asha1001 » Thu Aug 28, 2014 3:33 pm

A brokerage can sell you stocks for a very low commission $9.95 for 30 stocks in case of this company. However they can also give you a very bad spread and sell you a stock for say $9.25, when you could have bought it for $9.10. This is a hidden cost for you. Stocks remember don't have a single bid and ask, there are many entities in the market quoting bid and ask for the same stock. A brokerage decides where they route your order.

User avatar
JoMoney
Posts: 6081
Joined: Tue Jul 23, 2013 5:31 am

Re: Motif Investing - own the S&P500 for less

Post by JoMoney » Thu Aug 28, 2014 3:40 pm

berntson wrote:...Her strategy is to buy companies with the largest market cap first. She starts by saving for two months and then buys $10,000 in Apple stock. In the fourth month, Exxon's market cap is 74% of Apple's market cap, so Alice buys $7,400 worth of Exxon stock, enough for it to be worth 74% as much as her Apple holdings. (This leaves a little cash left over, but that's alright. She can invest it next month.) On the fifth month, Alice buys $5,600 worth of Microsoft stock. On the sixth month, she buys and $4,900 worth of Johnson and Johnson stock. And so on and so forth until she has bought all 500 stocks in the index.
Brilliant! :sharebeer
Now what's a strategy for unwinding the portfolio and maintaining the balance while selling off in a draw-down phase?
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham

inbox788
Posts: 5635
Joined: Thu Mar 15, 2012 5:24 pm

Re: Motif Investing - own the S&P500 for less

Post by inbox788 » Thu Aug 28, 2014 5:06 pm

asha1001 wrote:A brokerage can sell you stocks for a very low commission $9.95 for 30 stocks in case of this company. However they can also give you a very bad spread and sell you a stock for say $9.25, when you could have bought it for $9.10. This is a hidden cost for you. Stocks remember don't have a single bid and ask, there are many entities in the market quoting bid and ask for the same stock. A brokerage decides where they route your order.
Most major stocks (i.e. sp500) have penny spreads, so this shouldn't be much of a factor. Thinly traded stocks should probably be avoided. With 30 stocks in a motif, this effect is magnified up to 30X, so a penny here or there still isn't going to be a big deal, but 30 pink sheet stocks could wind up costing more than the commission.

Their clearing house seems typical, but why do they have limit orders? I though they did only market orders at this time. Is that a customer selection or does the brokerage alter the selection (to who's benefit)? And are Exchange Listed Options allowed in motifs?

http://public.s3.com/rule606/apex/APEX_ ... ule606.pdf

Buysider
Posts: 721
Joined: Sun May 18, 2008 9:36 pm

Re: Motif Investing - own the S&P500 for less

Post by Buysider » Fri Aug 29, 2014 5:31 am

Most major stocks (i.e. sp500) have penny spreads, so this shouldn't be much of a factor.
Yes, but the definition of a market order may vary between brokers. Are the traders routed immediately and executed at the best market price? Then, yes, we're talking a penny. How about the trade gets executed within a 15 minute window at a market best price within that window? A penny spread is dwarfed by the market motion and guess whose favor it goes to?

The SEC has had a few cases cracking down on this (ESPP, ISO, etc.), but I believe the practice still goes on. Don't know how it works with this firm.

YDNAL
Posts: 13774
Joined: Tue Apr 10, 2007 4:04 pm
Location: Biscayne Bay

Re: Motif Investing - own the S&P500 for less

Post by YDNAL » Fri Aug 29, 2014 5:59 am

Tanelorn [OP] » Wed Aug 27, 2014 12:23 am wrote:SPY, VOO, and Vanguard's S&P500 fund are very cheap at 0.05-0.10% ER, but could it be cheaper to just hold all the stocks yourself? Thanks to Motif Investing, a gimicky upstart brokerage where buying 30 stocks at a time only costs $9.95, this may be more practical than it has been for normal investors and especially those buying for the long term....
Are you an accumulator, Tanelorn?

For discussion, I'm going to suggest that a person is accumulating and this is 401K brokerage account -- where (s)he contributes $729.17 semi-monthly (maximum $17.5K).
  • 1. Is (s)he rebalancing this account with new money to execute 17 trades, 30 stocks each, $9.95 per trade, 24 times a year?
    2. "Cheaper and practical" ? :?
If the S&P 500 was part of my investment objective and plan, I would buy VOO and do other things with my time!... just a personal opinion.
Landy | Be yourself, everyone else is already taken -- Oscar Wilde

berntson
Posts: 1366
Joined: Mon Oct 29, 2012 12:10 pm

Re: Motif Investing - own the S&P500 for less

Post by berntson » Fri Aug 29, 2014 11:23 am

JoMoney wrote:
berntson wrote:...Her strategy is to buy companies with the largest market cap first. She starts by saving for two months and then buys $10,000 in Apple stock. In the fourth month, Exxon's market cap is 74% of Apple's market cap, so Alice buys $7,400 worth of Exxon stock, enough for it to be worth 74% as much as her Apple holdings. (This leaves a little cash left over, but that's alright. She can invest it next month.) On the fifth month, Alice buys $5,600 worth of Microsoft stock. On the sixth month, she buys and $4,900 worth of Johnson and Johnson stock. And so on and so forth until she has bought all 500 stocks in the index.
Brilliant! :sharebeer
Now what's a strategy for unwinding the portfolio and maintaining the balance while selling off in a draw-down phase?
I'll take that! :beer

Say Alice ends up with a $2 million portfolio and needs $60,000 a year to live on. She'll get most of that by just not reinvesting dividends (say dividends are $36,000). Now she needs $2,000 a month. She can just run the whole process in reverse. For the first six months (say) she sells Apple stock, meaning that she ends up selling about 20% of her total Apple holdings. Then she goes down the list after that. She sells Exxon so that her ratio of Exxon to Apple matches the total market, then Microsoft, then Johnson and Johnson... It will easily take her 20 years to work through the whole list. She will need to make no more than 600 sales over 20 years, resulting in average yearly fees of $30 (an expense ratio of 0.000015).

Alice could also start at the bottom of the list and work up if she likes large caps better than small caps. Or she could throw darts. Random selling would decrease tracking error relative to the S&P.

conlius
Posts: 92
Joined: Fri May 02, 2014 11:09 am

Re: Motif Investing - own the S&P500 for less

Post by conlius » Fri Aug 29, 2014 12:03 pm

Tax loss harvesting? You own the underlying stocks within a motif. You can sell an individual stock out of a motif if you don't want it anymore. In fact, they are working on a solution to automatically tax loss harvest all your current stocks with the click of a button.

Wash sale? If you have multiple motifs with overlapping stocks, that is your own fault. When you go to buy a motif, you have the option to customize and remove or replace any of your existing stocks for free.

A Motif is essentially building your own micro ETF. It gives you the benefit of tax loss harvesting individual stocks, re-balancing or replacing stocks, customizing other motifs to fit your approach, and the ability to buy/sell entire dollar values OR individual stocks. The product is extremely versatile in that you can pay the commission of a retail ETF buyer while also being the manager of the ETF's strategy.

This product is not a good replacement for passive index investing, primarily because of the $10 commission to buy/sell (there is also a $10 fee to re-balance and a $5 fee to sell an individual stock out of a motif). This product is more designed for someone who wants the ability to actively manage individual stock positions while retaining the ability to trade their entire position like an ETF.

freddie
Posts: 920
Joined: Sat Feb 08, 2014 11:06 pm

Re: Motif Investing - own the S&P500 for less

Post by freddie » Fri Aug 29, 2014 12:42 pm

inbox788 wrote:Taxwise, are motifs treated like a fund when you buy and sell? TLH and wash sales are complications or opportunities depending on accounting and tax treatment.

Let's say you bought a motif with 3 stocks ABC, and AB are winners while C is a loss. On the day you sell, you have a net loss on the motif, so if it's like a fund and you buy back the motif, it would be a wash sale, but if you just buy back stock C, could you still take the TLH on the motif? (similar to selling PPH for loss and buying JNJ).

So, out of your 17 motifs, take some of the top losers and identify the losing stocks from those motifs. Regroup the motifs so the losers are in one fund and shuffle the winners among other motifs. Buy back in same proportion as before you started. Individually the motifs are very different, but as a whole, you're keeping the same net position.

Using Bogleheads investment, let's say you bought VT in a year where US stocks went up, but ex-US went down a lot more, so VT is a loss. You sell VT and buy VTI and VXUS in similar allocation. Claim the tax loss on VT? What if the allocation is slightly or somewhat different? Only your accountant and IRS investigator know.

All these ideas are fairly theoretical questions since practically, it's going to be difficult to use motifs to replace low cost index funds. Only practical reason would be to lower costs and that means minimizing transactions, so that eliminates a lot of the potential benefits like TLH. Still, TLH is a two edge sword, since you're deferring taxes, not eliminating them. Those in the same tax bracket when investing vs retiring aren't impacted, and you're at risk for tax law changes. Tax policy changes like the 3.8% Medicare Tax may actually increase tax burden from TLH for some.
Tax wise there is no motif. Just a bunch of stock sales so in your example you would get a wash sale. You either end up overweighting (you buy more apple before you sell the stuff you want to unload), underweighting (you sell the apple and wait 30 days to buy it back), or trying to correllate (i.e. sell apple and buy google. Sounds a lot better when you say sell Exxon and buy Chevron). They all give you slightly different tracking risk. There are also some other fun things that you could do in theory to boost returns like shove the high dividend payers into a IRA.

I don't know how well this works for the S&P 500, but the REIT index for example has just under 40% of their holdings in 10 stocks. Buy 2 or 3 motifs will get you a huge chunk of the index so I expect your tracking error will be reasonable. I still expect vanguard is a lot more efficient at trading than I could be.

IPer
Posts: 1639
Joined: Sun Jul 28, 2013 8:51 pm

Re: Motif Investing - own the S&P500 for less

Post by IPer » Fri Aug 29, 2014 12:53 pm

I would have just said no but I need at least 5 characters!
Read the Wiki Wiki !

Tanelorn
Posts: 1546
Joined: Thu May 01, 2014 9:35 pm

Re: Motif Investing - own the S&P500 for less

Post by Tanelorn » Tue Sep 02, 2014 7:53 pm

Buysider wrote:
Much of the Motif cost savings comes from the flat fee with no annual charge, so you can view it as an illiquidity premium, small as it may be.
Investors focus on explicit costs because, well, they are explicit. Implicit, or hidden costs, are harder to compare and therefore are less easily analyzed. I think this thread is a little short on that.
Speaking of hidden costs, this thread discusses index front running and similar costs that could be easily avoided by an individual who holds the stocks directly (just don't bother changing your personal 500 index until a year after the "official" one does).

http://www.bogleheads.org/forum/viewtop ... 0&t=146013

It is estimated that this would save about 0.25% per year for the S&P500 and several times that for small caps.

Alex Frakt
Founder
Posts: 10853
Joined: Fri Feb 23, 2007 1:06 pm
Location: Chicago
Contact:

Re: Motif Investing - own the S&P500 for less

Post by Alex Frakt » Wed Sep 03, 2014 12:08 am

berntson wrote:Here is a similar strategy I've thought a bit about. Say that Alice is a new investor who plans to invest about $5,000 every month starting in January. She wants to hold individual stocks that more or less track the performance of the S&P. Her strategy is to buy companies with the largest market cap first. She starts by saving for two months and then buys $10,000 in Apple stock. In the fourth month, Exxon's market cap is 74% of Apple's market cap, so Alice buys $7,400 worth of Exxon stock, enough for it to be worth 74% as much as her Apple holdings. (This leaves a little cash left over, but that's alright. She can invest it next month.) On the fifth month, Alice buys $5,600 worth of Microsoft stock. On the sixth month, she buys and $4,900 worth of Johnson and Johnson stock. And so on and so forth until she has bought all 500 stocks in the index.
One problem. Historically the smaller cap components of the S&P500 have a higher return than the larger cap components. That's why the equal-weighted S&P 500 index has outperformed the cap-weighted index. You could start with the smallest constituents and work your way up, but then you can expect sizable tracking error.

The larger issue is if you accept the premise of indexing, why would you use the S&P 500 instead of TSM? The S&P 500 is a historical relic. Like the Dow 30 before it, it was only an approximation of the market, limited in scope by how quickly it could be calculated given the technology of the day.

Alex Frakt
Founder
Posts: 10853
Joined: Fri Feb 23, 2007 1:06 pm
Location: Chicago
Contact:

Re: Motif Investing - own the S&P500 for less

Post by Alex Frakt » Wed Sep 03, 2014 12:22 am

Buysider wrote:
Most major stocks (i.e. sp500) have penny spreads, so this shouldn't be much of a factor.
Yes, but the definition of a market order may vary between brokers. Are the traders routed immediately and executed at the best market price? Then, yes, we're talking a penny. How about the trade gets executed within a 15 minute window at a market best price within that window? A penny spread is dwarfed by the market motion and guess whose favor it goes to?
Also, even a penny a share spread at the brokerage is not insignificant given the incredibly low expenses we are talking about for Vanguard admiral funds/ETFs. Vanguard handles a great deal of its daily trading within its funds. There is no spread at all for these shares. And they are long-acknowledged masters at the transactional level when they do have go to the markets to buy or sell.

berntson
Posts: 1366
Joined: Mon Oct 29, 2012 12:10 pm

Re: Motif Investing - own the S&P500 for less

Post by berntson » Wed Sep 03, 2014 11:34 am

Alex Frakt wrote:
berntson wrote:Here is a similar strategy I've thought a bit about. Say that Alice is a new investor who plans to invest about $5,000 every month starting in January. She wants to hold individual stocks that more or less track the performance of the S&P. Her strategy is to buy companies with the largest market cap first. She starts by saving for two months and then buys $10,000 in Apple stock. In the fourth month, Exxon's market cap is 74% of Apple's market cap, so Alice buys $7,400 worth of Exxon stock, enough for it to be worth 74% as much as her Apple holdings. (This leaves a little cash left over, but that's alright. She can invest it next month.) On the fifth month, Alice buys $5,600 worth of Microsoft stock. On the sixth month, she buys and $4,900 worth of Johnson and Johnson stock. And so on and so forth until she has bought all 500 stocks in the index.
One problem. Historically the smaller cap components of the S&P500 have a higher return than the larger cap components. That's why the equal-weighted S&P 500 index has outperformed the cap-weighted index. You could start with the smallest constituents and work your way up, but then you can expect sizable tracking error.

The larger issue is if you accept the premise of indexing, why would you use the S&P 500 instead of TSM? The S&P 500 is a historical relic. Like the Dow 30 before it, it was only an approximation of the market, limited in scope by how quickly it could be calculated given the technology of the day.
This is a nice point. You could completely randomize the order. That would give the process a bit of a small-cap tilt. If you wanted to avoid the small-cap tilt, you could use weighted probabilities equivalent to a stock's market cap (i.e. the odds of picking Apple would be about 3.5% and the odds of picking Intel about 1% and so on). Using the cap-weighted probability approach, you could even sample the whole market instead of just the S&P 500.

Alex Frakt
Founder
Posts: 10853
Joined: Fri Feb 23, 2007 1:06 pm
Location: Chicago
Contact:

Re: Motif Investing - own the S&P500 for less

Post by Alex Frakt » Wed Sep 03, 2014 8:03 pm

berntson wrote:
Alex Frakt wrote:
berntson wrote:Here is a similar strategy I've thought a bit about. Say that Alice is a new investor who plans to invest about $5,000 every month starting in January. She wants to hold individual stocks that more or less track the performance of the S&P. Her strategy is to buy companies with the largest market cap first. She starts by saving for two months and then buys $10,000 in Apple stock. In the fourth month, Exxon's market cap is 74% of Apple's market cap, so Alice buys $7,400 worth of Exxon stock, enough for it to be worth 74% as much as her Apple holdings. (This leaves a little cash left over, but that's alright. She can invest it next month.) On the fifth month, Alice buys $5,600 worth of Microsoft stock. On the sixth month, she buys and $4,900 worth of Johnson and Johnson stock. And so on and so forth until she has bought all 500 stocks in the index.
One problem. Historically the smaller cap components of the S&P500 have a higher return than the larger cap components. That's why the equal-weighted S&P 500 index has outperformed the cap-weighted index. You could start with the smallest constituents and work your way up, but then you can expect sizable tracking error.

The larger issue is if you accept the premise of indexing, why would you use the S&P 500 instead of TSM? The S&P 500 is a historical relic. Like the Dow 30 before it, it was only an approximation of the market, limited in scope by how quickly it could be calculated given the technology of the day.
This is a nice point. You could completely randomize the order. That would give the process a bit of a small-cap tilt. If you wanted to avoid the small-cap tilt, you could use weighted probabilities equivalent to a stock's market cap (i.e. the odds of picking Apple would be about 3.5% and the odds of picking Intel about 1% and so on). Using the cap-weighted probability approach, you could even sample the whole market instead of just the S&P 500.
You are still missing the skewness of returns. Since a large part of index gains comes from a small number of stocks, the majority of samples will underperform the complete index.

inbox788
Posts: 5635
Joined: Thu Mar 15, 2012 5:24 pm

Re: Motif Investing - own the S&P500 for less

Post by inbox788 » Wed Sep 03, 2014 8:20 pm

Alex Frakt wrote:You are still missing the skewness of returns. Since a large part of index gains comes from a small number of stocks, the majority of samples will underperform the complete index.
Two things will reduce this effect. First, your sample of the index is small, so when you randomly hit one of these smaller high gain stocks, it will be a large part of your index subset, and overperform. On average, it will tend towards average. The Second thing is that long term, there is a reversion to the mean, so the longer you hold stocks, the less effect the outliers will have.

bpp
Posts: 1964
Joined: Mon Feb 26, 2007 12:35 pm
Location: Japan

Re: Motif Investing - own the S&P500 for less

Post by bpp » Wed Sep 03, 2014 11:46 pm

inbox788 wrote:
Alex Frakt wrote:You are still missing the skewness of returns. Since a large part of index gains comes from a small number of stocks, the majority of samples will underperform the complete index.
Two things will reduce this effect. First, your sample of the index is small, so when you randomly hit one of these smaller high gain stocks, it will be a large part of your index subset, and overperform. On average, it will tend towards average.
That doesn't help a single, sub-sampled portfolio. The problem with skewness is that any particular sub-sample has a better than 50% probability of underperforming the average.
The Second thing is that long term, there is a reversion to the mean, so the longer you hold stocks, the less effect the outliers will have.
I wouldn't hang much on RTM (I suspect it is mostly an artifact of back-fitting), but even if I did, I don't think it is meant by proponents to suggest that individual stocks have a tendency to return to the overall market average.

Alex Frakt
Founder
Posts: 10853
Joined: Fri Feb 23, 2007 1:06 pm
Location: Chicago
Contact:

Re: Motif Investing - own the S&P500 for less

Post by Alex Frakt » Wed Sep 03, 2014 11:58 pm

inbox788 wrote:
Alex Frakt wrote:You are still missing the skewness of returns. Since a large part of index gains comes from a small number of stocks, the majority of samples will underperform the complete index.
Two things will reduce this effect. First, your sample of the index is small, so when you randomly hit one of these smaller high gain stocks, it will be a large part of your index subset, and overperform. On average, it will tend towards average.
There are averages and there are averages. Yes, the mean return of a set of samples is expected to be the same as the return of the complete index. But the return of the median sample will be below the return of the complete index. In other words, the majority of investors who try to sample an index will underperform the index. It does not appear logical to me to accept a greater than 50% chance of underperforming an index fund in order to save a couple of basis points in expenses.
inbox788 wrote:The Second thing is that long term, there is a reversion to the mean, so the longer you hold stocks, the less effect the outliers will have.
RTM doesn't give you back gains you have missed by not holding a given individual stock. It does help you out if you have allocations to different asset classes and periodically rebalance, but that it far from what we are discussing here.

inbox788
Posts: 5635
Joined: Thu Mar 15, 2012 5:24 pm

Re: Motif Investing - own the S&P500 for less

Post by inbox788 » Thu Sep 04, 2014 10:49 am

Alex Frakt wrote:There are averages and there are averages. Yes, the mean return of a set of samples is expected to be the same as the return of the complete index. But the return of the median sample will be below the return of the complete index. In other words, the majority of investors who try to sample an index will underperform the index. It does not appear logical to me to accept a greater than 50% chance of underperforming an index fund in order to save a couple of basis points in expenses.
Thanks for pointing out the difference between mean and median return. I'll have to consider the the effects are going to be significant or how many samples/periods will be needed to minimize their effect so it's not significant.

Yes, while in a given interval, "the majority of investors who try to sample an index will underperform the index. ", those few that outperform one period will underperform in others, while others that have underperformed will outperform. So over multiple periods, everyone will approach the mean, which is what I mean (no pun intended) by reversion to the mean. Over many samples and many periods, everyone should approach index, most from below, few from above (per your skewness).

I may not understand this skewness correctly, but I'm imagining a fair (i.e. no zeros) roulette table. Someone betting on a single number faces 1:35 odds. So yes, those that bet on a single number will likely lose and a few lucky ones that hit their number will come out way ahead, but no money enters or leaves the table on average. If someone covered every number, they would be assured 0 return, the return of the SP500. Now if everyone played tens or hundreds of times, and didn't just cover 1 number, but 5 or 20 numbers, they would all approach the 0 return.

Alex Frakt
Founder
Posts: 10853
Joined: Fri Feb 23, 2007 1:06 pm
Location: Chicago
Contact:

Re: Motif Investing - own the S&P500 for less

Post by Alex Frakt » Fri Sep 05, 2014 12:13 pm

inbox788 wrote:I may not understand this skewness correctly, but I'm imagining a fair (i.e. no zeros) roulette table. Someone betting on a single number faces 1:35 odds. So yes, those that bet on a single number will likely lose and a few lucky ones that hit their number will come out way ahead, but no money enters or leaves the table on average. If someone covered every number, they would be assured 0 return, the return of the SP500. Now if everyone played tens or hundreds of times, and didn't just cover 1 number, but 5 or 20 numbers, they would all approach the 0 return.
I see the problem. Skewness refers to outcomes that are not normally distributed. Think of a game in which you get to select as many cards as you want from a face-down deck. It costs $10 for each card you pick up and only the clubs pay off. Let's also imagine this is a benevolent casino that provides a 10% return on your investment. So you get $44 if you get clubs and $0 for anything else. The index return of this game (buying all of the cards) is 10%. But most of the people who buy fewer cards will receive less than that.

Stock market returns are even more oddly distributed than this, you have a small cluster that goes to 0, a large cluster that's centered a bit under the mean and a very small number that goes up 20x or more. The returns on any sample depend on whether you are lucky enough to get one of the high-fliers.

Tanelorn
Posts: 1546
Joined: Thu May 01, 2014 9:35 pm

Re: Motif Investing - own the S&P500 for less

Post by Tanelorn » Fri Sep 05, 2014 12:37 pm

Alex Frakt wrote:Think of a game in which you get to select as many cards as you want from a face-down deck. It costs $10 for each card you pick up and only the clubs pay off. Let's also imagine this is a benevolent casino that provides a 10% return on your investment. So you get $44 if you get clubs and $0 for anything else. The index return of this game (buying all of the cards) is 10%. But most of the people who buy fewer cards will receive less than that.
This model makes it easy to see the TLH benefits when outcomes are skewed. Let's say you're in a 50% bracket for simplicity - now you only pay $5 net when you lose, but you keep the whole $44 tax deferred when you win. It makes it easy to keep playing until you hit a winner, since your effective cost is half as much.

Instead of putting in $40 to make $44, now you put in $40, get $15 worth of tax benefits on your $30 of losses (net $25 cost to play 4 cards), and get to keep your one winner for $44 deferred. Yes, the winner has a $34 unrealized gain ($17 tax liability, $17 profit, $10 basis), which means if you sold it you'd end up with $42, having paid half your nets $4 gain in taxes ($42 = $10 basis + $17 net profit on winner + $15 tax benefits for losses).

berntson
Posts: 1366
Joined: Mon Oct 29, 2012 12:10 pm

Re: Motif Investing - own the S&P500 for less

Post by berntson » Fri Sep 05, 2014 12:59 pm

Alex Frakt wrote:
inbox788 wrote:I may not understand this skewness correctly, but I'm imagining a fair (i.e. no zeros) roulette table. Someone betting on a single number faces 1:35 odds. So yes, those that bet on a single number will likely lose and a few lucky ones that hit their number will come out way ahead, but no money enters or leaves the table on average. If someone covered every number, they would be assured 0 return, the return of the SP500. Now if everyone played tens or hundreds of times, and didn't just cover 1 number, but 5 or 20 numbers, they would all approach the 0 return.
I see the problem. Skewness refers to outcomes that are not normally distributed. Think of a game in which you get to select as many cards as you want from a face-down deck. It costs $10 for each card you pick up and only the clubs pay off. Let's also imagine this is a benevolent casino that provides a 10% return on your investment. So you get $44 if you get clubs and $0 for anything else. The index return of this game (buying all of the cards) is 10%. But most of the people who buy fewer cards will receive less than that.

Stock market returns are even more oddly distributed than this, you have a small cluster that goes to 0, a large cluster that's centered a bit under the mean and a very small number that goes up 20x or more. The returns on any sample depend on whether you are lucky enough to get one of the high-fliers.
Inbox' original example is one in which returns are not evenly distributed, at least on each individual spin of the wheel. The mean return (zero) is higher than the median (negative). The point that inbox was trying to make is that if you play a game with a non-normal distribution enough times, the overall distribution has less skew and lower keratosis (everyone gets closer and closer to the average). So while most S&P samplers may get subpar returns when looking at short periods, it may be that most get the average over long periods.

The point about skew is a nice point Alex. It would be interesting to figure how it would have actually influenced returns over long timer periods using historical data.

Alex Frakt
Founder
Posts: 10853
Joined: Fri Feb 23, 2007 1:06 pm
Location: Chicago
Contact:

Re: Motif Investing - own the S&P500 for less

Post by Alex Frakt » Fri Sep 05, 2014 1:40 pm

berntson wrote:The point that inbox was trying to make is that if you play a game with a non-normal distribution enough times, the overall distribution has less skew and lower keratosis (everyone gets closer and closer to the average). So while most S&P samplers may get subpar returns when looking at short periods, it may be that most get the average over long periods.
Gambler's fallacy. The expected return of each new period is the same as the index return. But you cannot expect to be compensated for the high-fliers you missed previously. If you sample, you will have winners and losers (relative to the index return). The way stock returns are distributed, you will have more losers than winners. Even though the winners win more than the losers lose, I still don't think it's a bet worth taking when the alternative is so inexpensive.

Tanelorn
Posts: 1546
Joined: Thu May 01, 2014 9:35 pm

Re: Motif Investing - own the S&P500 for less

Post by Tanelorn » Fri Sep 05, 2014 1:54 pm

Alex Frakt wrote:
berntson wrote:The point that inbox was trying to make is that if you play a game with a non-normal distribution enough times, the overall distribution has less skew and lower keratosis (everyone gets closer and closer to the average). So while most S&P samplers may get subpar returns when looking at short periods, it may be that most get the average over long periods.
Gambler's fallacy. The expected return of each new period is the same as the index return. But you cannot expect to be compensated for the high-fliers you missed previously. If you sample, you will have winners and losers (relative to the index return). The way stock returns are distributed, you will have more losers than winners. Even though the winners win more than the losers lose, I still don't think it's a bet worth taking when the alternative is so inexpensive.
I agree with berntson. He's not saying that the losers are "due" for a winner just because they lost, but rather that a sample of a large of number of trials is likely to be close to the mean. In this case, each stock for each time period is one (independent?) trial with the same expected return. Yes, when you sample some people will underperform and fewer will outperform and by a bigger amount than the losers (due to skewness), but as they own more stocks and for a longer period of time, they are more and more likely to achieve a return close to the mean. This is just the central limit theorem at work.

http://en.m.wikipedia.org/wiki/Central_limit_theorem

berntson
Posts: 1366
Joined: Mon Oct 29, 2012 12:10 pm

Re: Motif Investing - own the S&P500 for less

Post by berntson » Sat Sep 06, 2014 5:31 pm

Alex Frakt wrote:
berntson wrote:The point that inbox was trying to make is that if you play a game with a non-normal distribution enough times, the overall distribution has less skew and lower keratosis (everyone gets closer and closer to the average). So while most S&P samplers may get subpar returns when looking at short periods, it may be that most get the average over long periods.
Gambler's fallacy. The expected return of each new period is the same as the index return. But you cannot expect to be compensated for the high-fliers you missed previously. If you sample, you will have winners and losers (relative to the index return). The way stock returns are distributed, you will have more losers than winners. Even though the winners win more than the losers lose, I still don't think it's a bet worth taking when the alternative is so inexpensive.
It's not the gambler's fallacy. I was thinking it was just standard regression to the mean, a theorem from statistics. Say you and I play a coin flipping game where the loser pays the winner a dollar after each flip. We flip the coin once and you win. I pay you a dollar. Of course, this has no effect on future flips. But as we perform more and more flips, your expected advantage asymptotically approaches zero.

But maybe the compounding makes a difference. Suppose by chance your concentrated portfolio returns double the total market and mine gets only market returns. If we both expect to get market returns going forward, we can expect your portfolio to always be double my portfolio.

Tanelorn
Posts: 1546
Joined: Thu May 01, 2014 9:35 pm

Re: Motif Investing - own the S&P500 for less

Post by Tanelorn » Thu Sep 11, 2014 11:30 am

Interview with Motif Investing's founder about his goals and the company.

http://www.reddit.com/r/IAmA/comments/2 ... man_sachs/

Sounds like there are lots of things in the pipeline but he was cagey on some of the details of how the business works and is growing.

User avatar
marklearnsbogle
Posts: 133
Joined: Tue Jul 15, 2014 7:56 pm
Location: New York City

Re: Motif Investing - own the S&P500 for less

Post by marklearnsbogle » Sat Nov 01, 2014 12:15 pm

JoMoney wrote:Also, the actual realized expenses of Vanguard's Admiral class 500 index VFIAX is surprisingly lower than you estimate. The funds internal brokerage/trading expenses are extremely low (much less than 1 basis point if you look at the what's stated in the SAI relative to the funds assets).
If you compare it to the actual index it's also very apparent just how low the overall cost is:
Image
Over the past 14 years the S&P500 Index return was $17,411.53
VFIAX return was 17,361.67 the effective difference/CAGR is .02%
IShares IVV return was $17,253.98 the effective difference/CAGR is .06%

Over the 5 year period from 2008-2013 VFIAX actually BEAT the index:
Image
Hi JoMoney - I was going to post this question, but then found this thread. There's something I do not know how to figure out math-wise, and perhaps it is complicated (or deceptively easy). If considering exchanging TSM Admiral class for VG 500 Admiral class, one would come out with fewer shares since the price per share of the 500 is much more than of TSM. How does one figure whether it is more advantageous to own more shares at a lower price (TSM) or fewer shares at a higher price (500)? All of this is in a Roth with divs reinvesting for the time being. Thanks!
"Nothing is simpler than owning the stock market and holding it forever, and that’s essentially the idea behind the index fund.” - Bogle.

User avatar
ogd
Posts: 4854
Joined: Thu Jun 14, 2012 11:43 pm

Re: Motif Investing - own the S&P500 for less

Post by ogd » Sat Nov 01, 2014 12:38 pm

marklearnsbogle wrote:How does one figure whether it is more advantageous to own more shares at a lower price (TSM) or fewer shares at a higher price (500)?
There is no difference. The # of component stocks (e.g. Apple) per dollar invested will be the roughly the same (500 will have a little more because it's large cap oriented).

If you invest in ETFs you're limited to whole numbers of shares, so larger unit prices create a bit of an annoyance because of leftover cash. Still, a very small deal, and no such problem with mutual funds.

Post Reply