Jim Cramer doesn’t beat the market

[Ad Orientem]

I don't have a firm rebalancing point for my VP. I stick to the 15/35 bands for the HBPP. For my VP I follow my gut when weighing the economic/political climate. That said I don't think I would be comfortable if either of the two components reached a 2:1 ratio relative to the other. Nor would I be comfortable if my VP reached parity with my HBPP. If that happened I'd likely hit the reset button. But I really like Jack Bogle's advice on rebalancing. "Rebalance when you think you need to, not based on a calendar or set of numbers." If your portfolio asset allocation is keeping you awake at night, then you probably should rebalance.

[Kbg]

Without going into a ton of detail, market cap weighted indexes have a lot of problems, with a couple of advantages. If you want to beat the S&P 500 over the long haul just buy and hold RSP. The only time it doesn't beat the SPY is at the end of expensive bull markets. It will spank the SPY coming out of a bear as well as most of the way up. ( and it's an OK timing signal when compared with SPY...basically if it is doing better over x period than spy it is a good time to be in stocks, when it is doing worse there is danger in the market )

[MachineGhost]

Without going into a ton of detail, market cap weighted indexes have a lot of problems, with a couple of advantages. If you want to beat the S&P 500 over the long haul just buy and hold RSP. The only time it doesn't beat the SPY is at the end of expensive bull markets. It will spank the SPY coming out of a bear as well as most of the way up. ( and it's an OK timing signal when compared with SPY...basically if it is doing better over x period than spy it is a good time to be in stocks, when it is doing worse there is danger in the market )

It's just the size effect. It will get creamed in a bear more so than big bro. Perhaps theres an effective switching strategy?

Incidentally, RSP's been bearish for the last 10 months but I just got a buy signal on Schwab Small Cap the other day. Everything in my breadth portfolio is now bullish except for NASDAQ COMP, S&P 600, Materials and Financial. That sure looks like the dumb money is still not participating.

[Kbg]

Without going into a ton of detail, market cap weighted indexes have a lot of problems, with a couple of advantages. If you want to beat the S&P 500 over the long haul just buy and hold RSP. The only time it doesn't beat the SPY is at the end of expensive bull markets. It will spank the SPY coming out of a bear as well as most of the way up. ( and it's an OK timing signal when compared with SPY...basically if it is doing better over x period than spy it is a good time to be in stocks, when it is doing worse there is danger in the market )

It's just the size effect. It will get creamed in a bear more so than big bro. Perhaps theres an effective switching strategy?

Incidentally, RSP's been bearish for the last 10 months but I just got a buy signal on Schwab Small Cap the other day. Everything in my breadth portfolio is now bullish except for NASDAQ COMP, S&P 600, Materials and Financial. That sure looks like the dumb money is still not participating.

I don't disagree that some of it is the size effect. This is clearly the case. There's more to it than that though. The mathematics of scale normally make a cap weighted index under perform the mean of its constituents over the long haul. Companies that are largest have the biggest impact on the index and when the largest companies begin to shrink (as they inevitably will) they have a disproportionate negative impact. It is also highly likely they can't match the growth rates of smaller companies. So forget RSP, pick any random 50 stocks in the S&P 500 and you will likely beat the S&P 500 over time. (theory not necessarily reality...reality means trading costs, discipline, few/no trading errors etc.) Obviously there are times that the biggest companies perform the best and this can last for quite a while. However, over very long periods of time the math is decidedly against you.

Since RSP's creation (May 2003) it has outperformed SPY 10.78 CAGR/-59.92 vs 8.64/-55.17. Additionally, RSP has beat SPY in 10 out of 14 years including the partial years of 2003 and 2016 to date. If we use SPXEW vs. SPX we can extend back to 1/1/90 and the stats are 8.62/-60.81 vs. 6.93/-56.77. Meanwhile, replacing SPY with RSP in the PP since 2005 bumps the CAGR up to 8.21 vs 7.81 while adding slightly over 1% in additional max drawdown.

For the PP I personally believe RSP is a better choice than SPY if one isn't going to diversify the stock component explicitly to small caps. A 2%+ (live ETFs/13 years) and 1.5%+ (Indexes/36 years) CAGR advantage over a long period of time including both major bull and bear markets is pretty compelling evidence to me. The sharpe is better for RSP/SPXEW as well .39/.29 vs. .30/.20...so by the standard definition it's also a better risk adjusted investment. Additionally, within the context of a PP the time when RSP will be not outperforming SPY is exactly when LTTs and Gold are likely stepping up so why not juice the stock component slightly for the times when stocks are outperforming?

I've not found a timing mechanism that beats a buy and hold of RSP vs. SPY after costs. Not saying it can't be done, I just haven't found one nor put a ton of time into it. Not including costs, very short term momentum measures will beat an RSP buy and hold. I do monitor the ratio of RSP to SPY however as one of several market indicators I use. When RSP is outperforming SPY the stock market is usually doing pretty well internally.

[MachineGhost]

Hmm, that's very interesting! If what you say is true about equal weighting being superior for compounding, then that explains why I could not see any edge over the long-term by focusing on any of the Mega to Micro size categories (they all returned roughly the same). It may also explain why the Wilshire 5000 Equal Weight is so bloody seductive. I had chalked it up to the uninvestable bottom 10% of the bottom 2%.

Since 2003 I get 8.26% and -16.35% using RSP PP vs 7.67% and -15.24% for SPY PP. And 11.75% CAGR and -22.87% using RSP vs 10.81% and -18.56% using SPY and both using trend following.

So is using RSP superior to the Golden Butterfly?

EDIT: I just read Tyler's info about the Golden Butterfly and it seems it has less to do with the SCV and more to do with the strategic allocation. I can't help but think the weights are a data mining artifact because I can't think of a logical basis for why 40/20/20/20 makes sense. Still, it beats the PP on return, MaxDD and time underwater and that's rather unusual!

EDIT EDIT: SCV plays a key role in halving the time underwater. Doing 10%/20%/10% into large, mid and small blends also has the same effect. Still seems to me like its beneficial to get equal market cap exposure, but I don't wanna go down that rabbit hole again. Tyler has size and style time series data that he hasn't put up yet and that should help tell the tale of what is what.

Platinum Butterfly:


Golden Butterfly:


Platinum PP:


Golden PP:


Tired Old PP:

[Mr Vacuum]

It is interesting how similar the effect is across RSP PP, Golden PP, and Platinum PP. (I'd like to see Tyler's lost money and made the average numbers for RSP.) The appeal of RSP is it's just the one symbol to trade. Costs are a little higher, though: 0.40% expense ratio vs. <0.10% for VTI/VB. Tax cost ratio looks ok on morningstar.

[Kbg]

It is interesting how similar the effect is across RSP PP, Golden PP, and Platinum PP. (I'd like to see Tyler's lost money and made the average numbers for RSP.) The appeal of RSP is it's just the one symbol to trade. Costs are a little higher, though: 0.40% expense ratio vs. <0.10% for VTI/VB. Tax cost ratio looks ok on morningstar.

Costs are certain outperformance is not. However, figures provided include costs. So the outperformance is net not gross. If I recall both the standard broad etfs and RSP have reduced costs. If one was really diligent they could track down historical fees and compare to today's cost spreads. If the spread is greater assume the outperformance will be less and vice versa.

[Kbg]

One important thing I forgot in my first RSP post; if one believes in return to mean or buy low sell high for stocks RSP/equal weight does that. Cap weight does the opposite.

[MachineGhost]

It is interesting how similar the effect is across RSP PP, Golden PP, and Platinum PP. (I'd like to see Tyler's lost money and made the average numbers for RSP.) The appeal of RSP is it's just the one symbol to trade. Costs are a little higher, though: 0.40% expense ratio vs. <0.10% for VTI/VB. Tax cost ratio looks ok on morningstar.

RSP isn't equal weighted on the size exposure though. PortfolioCharts is all value weight and even if you equalize the size exposure, it won't be that in actuality.

I noticed last night there are equal weight sector ETF's available. An equal weight portfolio of those could be very, very interesting! It might blow RSP out of the water.

Of course, my dream is still to replicate the non-investable Wilshire 5000 Equal Weight which returned 17% a year since the early 70's.

There are three potential sources of excess return for an equal-weighted strategy. One is the small size tilt that results from underweighting mega-cap stocks and overweighting smaller stocks. The second is a value tilt that results from underweighting overpriced glamour stocks. The third potential source of excess return is from the contrarian rebalancing. In order to maintain equal weightings, the index must sell stocks that have recently appreciated and buy stocks that have recently declined. We have back-tested performance on the S&P 500 Equal Weight Index going back to 1991. Of the 70 basis-point average monthly return in the equal-weighted index, 55 basis points came from pure market exposure. The equal- weighted index did beat the market-cap-weighted index by about 19 basis points per month over the past 20 years. To examine the source of this outperformance requires a deeper dive into the numbers.

Compared to the market-cap-weighted S&P 500, equal weighting has a greater exposure (beta) to small-cap stocks and value stocks but a much lower exposure to momentum. The lower exposure to momentum is a result of the forced rebalance strategy where past winners are sold off in favor of past losers--the exact opposite of a momentum strategy. Of the 19 basis points of outperformance, we found that the tilt toward small size contributed 6 basis points, the value tilt contributed 9 basis points, while the bet against momentum actually detracted 7 basis points. This leaves about 11 basis points of alpha. A paper by Yuliya Plyakha, Raman Uppal, and Grigory Vilkov demonstrates that the source of the alpha in an equal-weighted strategy is the rebalance.

A monthly alpha of 11 basis points would be great if it could be obtained in practice. Conducting the same analysis on RSP since its inception in mid-2003 shows an alpha of 7 basis points per month. What happened to the 11 basis points we found in backtest? Most of it was likely eaten by trading costs and RSP's 0.40% expense ratio. While the modest outperformance of RSP is nice, it is important to remember that most of RSP's return can be attributed to beta exposures.

http://ibd.morningstar.com/article/arti ... ,%20brf295

So .84% edge per year? Pretty flimsy. But you could probably deal with the negative momentum alpha by using trend following.

I suspect the reason the GB improves things is its putting pseudo-momentum at the top and value at the bottom.

[Tyler]

RSP isn't equal weighted on the size exposure though. PortfolioCharts is all value weight and even if you equalize the size exposure, it won't be that in actuality.

True. The thing I like about the combination of half large cap fund and half small cap fund (for the stock portion of the GB portfolio) is that the resulting mix is a decent approximation for equal weights across large/mid/small caps. It's certainly not a true equal weight for every individual company, but it also covers many more companies (including small caps) than RSP.

Tyler has size and style time series data that he hasn't put up yet and that should help tell the tale of what is what.

Yep -- I'll be updating the data on the site soon (adding MCV and MCG and cleaning up a few older series with better data). I'm currently working with the Simba guys to update the master spreadsheet with my new data, and the site tools will follow after that.

[MachineGhost]

True. The thing I like about the combination of half large cap fund and half small cap fund (for the stock portion of the GB portfolio) is that the resulting mix is a decent approximation for equal weights across large/mid/small caps. It's certainly not a true equal weight for every individual company, but it also covers many more companies (including small caps) than RSP.

Yes, but that turned out to be a red herring. It is the multi-factor value, concentrated into small, that really drives the GB not the pseudo-equal weighting. Since people can't stick with value investing long-term due to behaviorial reasons, it will continue to endure. All of the market-beating gurus are more or less bottoms up value investors.

Now personally, I think restricting value investing to just small size is probably leaving money on the table, but its just a hypothesis without any empirical support until it is compared to Small Cap Value. And all value may not work in the PP because it relies on beta/momentum to hit rebalancing bands. I'm thinking about taking a flyer on it, though. I got enough other Prosperity exposure that I'm not reliant on stocks alone to deliver the full 25% impact.

[Kbg]

For my VP PP that I post on in the VP section I use SPXL and TNA 50/50 for the stock component. My rationale is very simple for this. As a general rule large cap performs better late stage bull and in bear markets and small cap performs better early stage bull markets.