Engineering Targeted Returns and Risk

[MachineGhost]

Here are the latest backtest results.  Everything is rebalanced every three months and TrendFollow is rebalanced monthly.  No transaction costs or taxes have been deducted.

GSD = Geometric Standard Deviation
MG = Custom risk metric I'm working on.
Ra = Risk Adjusted to HBPP

Code: Select all

Algorithm Quarterly Rebalanced          	Start      CAR     MaxDD   Score
Equal Weight, 10% Target GSD 60d, AbsMoMo	11/02/70   8.74% -10.26%    2.37
GSD Weight 60d, 1% Target MG 60d, AbsMoMo	11/02/70   5.77%  -4.27%    2.31
Equal Weight, 1% Target MG 60d, TrendFollow	11/02/70   5.74%  -5.34%    2.24
GSD Weight 60d, 10% Target GSD 60d, AbsMoMo	11/02/70   8.45%  -8.55%    2.23
Equal Weight, 1% Target MG 60d, AbsMoMo 	11/02/70   5.72%  -4.95%    2.21
MaxDD Weight, 1% Target MG 60d, TrendFollow	11/02/70   5.72%  -5.35%    2.12
GSD Weight 60d, 10% Target MaxDD, AbsMoMo	11/02/70   5.21%  -4.42%    2.06
Equal Weight, 10% MaxDD, AbsMoMo        	11/02/70   5.43%  -7.56%    2.00
Equal Weight, 10% Target GSD 60d, TrendFollow	11/02/70   8.47%  -9.57%    1.99
Equal Weight, 10% Target MaxDD, TrendFollow	11/02/70   5.37%  -7.49%    1.94
MG Weight All, 1% Target MG 60d, AbsMoMo	11/02/70   5.90%  -5.35%    1.91
MG Weight All, 10% Target GSD 60d, AbsMoMo	11/02/70   8.60% -11.59%    1.79
Equal Weight, AbsMoMo                   	11/02/70   8.42% -16.59%    1.78
GSD Weight 60d, 1% Target MG 60d, TrendFollow	11/02/70   5.44%  -5.58%    1.77
MaxDD Weight, 10% Target MaxDD, TrendFollow	11/02/70   5.39%  -6.41%    1.76
GSD Weight 60d, 10% Target MaxDD, TrendFollow	11/02/70   5.05%  -4.90%    1.70
GSD Weight 60d, AbsMoMo                 	11/02/70   9.05% -15.87%    1.68
MaxDD Weight, 10% Target GSD 60d, TrendFollow	11/02/70   7.76% -10.16%    1.66
Equal Weight, TrendFollow               	11/02/70   8.16%  -7.38%    1.62
MG Weight All, 1% Target MG 60d         	11/02/70   5.89%  -7.88%    1.62
RaGSD Weight, 1% Target MG 60d          	11/02/70   5.60%  -7.50%    1.61
Equal Weight, 1% Target MG 60d          	11/02/70   5.58%  -7.49%    1.60
RaMG Weight, 1% Target MG 60d           	11/02/70   5.58%  -7.51%    1.60
MaxDD Weight, 1% Target MG 60d          	11/02/70   5.69%  -6.62%    1.59
MG Weight 60d, 10% Target GSD 60d, AbsMoMo	11/02/70   8.66% -15.93%    1.57
MG Weight All, 10% MaxDD, AbsMoMo       	11/02/70   5.64%  -7.42%    1.56
MaxDD Weight, AbsMoMo                   	11/02/70   9.12% -16.73%    1.50
GSD Weight 60d, TrendFollow             	11/02/70   8.59% -15.16%    1.46
GSD Weight 60d, 10% Target GSD 60d, TrendFollow	11/02/70   7.25%  -9.62%    1.42
MaxDD Weight, TrendFollow               	11/02/70   9.00% -15.98%    1.41
DynamicMG Weight, 1% Target RaMG 60d    	11/02/70   6.65% -12.68%    1.36
RaGSD Weight, 10% Target GSD 60d        	11/02/70   8.03% -15.49%    1.35
RaMG Weight, 10% Target GSD 60d         	11/02/70   8.04% -15.50%    1.34
MG Weight All, 1% Target MG All, AbsMoMo	11/02/70   4.55%  -4.33%    1.34
Equal Weight, 10% Target MaxDD          	11/02/70   5.24% -10.32%    1.32
GSD Weight 60d, 1% Target MG 60d        	11/02/70   5.18%  -5.42%    1.28
Equal Weight, 10% Target GSD 60d        	11/02/70   7.81% -15.49%    1.27
Equal Weight, 1% Target MG All, AbsMoMo 	11/02/70   4.40%  -4.35%    1.23
MG Weight All, 10% Target GSD 60d       	11/02/70   8.42% -15.03%    1.23
MaxDD Weight, 10% Target GSD 60d        	11/02/70   8.08% -14.25%    1.20
MG Weight 60d, AbsMoMo                  	11/02/70   9.20% -18.84%    1.17
MG Weight All, AbsMoMo                  	11/02/70   9.44% -20.56%    1.16
DynamicMG Weight, 1% Target MG All      	11/02/70   7.03% -12.68%    1.14
GSD Weight 60d, 1% Target MG All, AbsMoMo	11/02/70   4.30%  -4.04%    1.13
DynamicGSD Weight, 1% Target RaMG 60d   	11/02/70   5.12% -12.68%    1.12
MG Weight 60d, 1% Target MG 60d, AbsMoMo	11/02/70   6.52% -13.20%    1.09
DynamicMG Weight, 1% Target MG 60d      	11/02/70   9.91% -21.12%    1.07
RaGSD Weight, 10% Target MaxDD          	11/02/70   5.25% -10.35%    1.07
RaMG Weight, 10% Target MaxDD           	11/02/70   5.23% -10.35%    1.06
DynamicGSD Weight, 1% Target MG 60d     	11/02/70   7.75% -20.33%    1.05
DynamicGSD Weight, 1% Target MG All     	11/02/70   5.51% -12.68%    1.03
DynamicMG Weight, 10% Target GSD 60d    	11/02/70  10.96% -22.78%    1.03
MG Weight All, 10% Target MaxDD         	11/02/70   5.58% -10.61%    0.99
MG Weight 60d, 1% Target MG 60d         	11/02/70   7.31% -16.82%    0.97
MG Weight 60d, 10% Target GSD 60d, TrendFollow	11/02/70   7.46% -13.98%    0.96
DynamicGSD Weight, 10% Target GSD 60d   	11/02/70   9.12% -19.39%    0.94
MG Weight 60d, TrendFollow              	11/02/70   8.93% -21.22%    0.94
MG Weight 60d, 1% Target RaMG 60d       	11/02/70   5.42% -12.50%    0.90
MaxDD Weight, 10% Target MaxDD          	11/02/70   5.25%  -9.63%    0.88
MG Weight 60d, 1% Target MG 60d, TrendFollow	11/02/70   5.59% -11.50%    0.87
MG Weight 60d, 1% Target MG All         	11/02/70   5.64% -12.50%    0.87
MG Weight All, 1% Target MG All         	11/02/70   4.51%  -6.02%    0.85
MG Weight 60d, 10% Target GSD 60d       	11/02/70   9.00% -24.66%    0.84
MG Weight 60d, 10% Target MaxDD, TrendFollow	11/02/70   5.27% -10.78%    0.84
MG Weight 60d, 10% Target MaxDD, AbsMoMo	11/02/70   5.40% -11.74%    0.82
DynamicGSD Weight, 10% Target MaxDD     	11/02/70   6.23% -16.59%    0.80
Equal Weight                            	11/02/70   7.77% -25.17%    0.80
DynamicMG Weight, 10% Target MaxDD      	11/02/70   7.72% -18.19%    0.79
RaGSD Weight, 1% Target MG All          	11/02/70   4.33%  -5.54%    0.78
Equal Weight, 1% Target MG All          	11/02/70   4.32%  -5.53%    0.77
RaMG Weight, 1% Target MG All           	11/02/70   4.32%  -5.54%    0.77
GSD Weight 60d, 10% Target MaxDD        	11/02/70   4.71%  -7.42%    0.76
MG Weight All                           	11/02/70   9.24% -33.25%    0.76
MaxDD Weight                            	11/02/70   8.55% -32.11%    0.75
MG Weight 60d, 10% Target MaxDD         	11/02/70   6.20% -13.88%    0.74
RaGSD Weight 60d                        	11/02/70   8.97% -35.21%    0.71
RaMG Weight 60d                         	11/02/70   8.96% -35.71%    0.70
MaxDD Weight, 1% Target MG All          	11/02/70   4.34%  -6.01%    0.68
DynamicMG Weight 60d                    	11/02/70  10.45% -43.16%    0.59
GSD Weight 60d                          	11/02/70   7.17% -28.20%    0.59
DynamicGSD Weight 60d                   	11/02/70  10.31% -43.16%    0.58
MG Weight 60d                           	11/02/70   7.62% -33.06%    0.52
GSD Weight 60d, 1% Target MG All        	11/02/70   4.12%  -5.13%    0.50
GSD Weight 60d, 10% Target GSD 60d      	11/02/70   5.96% -19.64%    0.43
MG Weight 60d, 1% Target MG All, AbsMoMo	11/02/70   4.52% -11.74%    0.43
Equal Weight, 1% Target RaMG 60d, AbsMoMo	11/02/70   3.99%  -4.26%    0.39
MG Weight All, 1% Target RaMG 60d, AbsMoMo	11/02/70   3.98%  -4.28%    0.37
RaGSD Weight, 1% Target RaMG 60d        	11/02/70   3.98%  -4.45%    0.34
Equal Weight, 1% Target RaMG 60d        	11/02/70   3.97%  -4.45%    0.33
RaMG Weight, 1% Target RaMG 60d         	11/02/70   3.97%  -4.44%    0.33
MG Weight All, 1% Target RaMG 60d       	11/02/70   3.98%  -4.64%    0.33
GSD Weight 60d, 1% Target RaMG 60d, AbsMoMo	11/02/70   3.95%  -4.21%    0.31
MaxDD Weight, 1% Target RaMG 60d        	11/02/70   3.95%  -4.67%    0.27
MG Weight 60d, 1% Target RaMG 60d, AbsMoMo	11/02/70   4.15% -11.74%    0.20
GSD Weight 60d, 1% Target RaMG 60d      	11/02/70   3.91%  -4.51%    0.17

[Stefan]

MG: I am not clear how you scored them, but this one looks the best:
GSD Weight 60d, 10% Target GSD 60d, AbsMoMo 11/02/70  8.45%  -8.55% 

Based on your abreviations, it looks like the approach suggested here:

To synthesize what I already said in this thread, there are 2 forms of risk management to be overlayed on top of PP (or AWP) to make the system more palatable to the conservative investor:

1.  Risk Moderation  achievable by :
            a.    Using a risk parity allocation formula based on volatility weights
            b.    Managing the whole portfolio volatility to a target.

2.  Risk Containment achievable by:
          a.   Asset classes trend. Capturing most of the trend upside and avoiding most of the trend downside.
          b.   Minimizing exposure gradually in times of market turbulence.

It would be interesting to see the equity curve & drawdown over the HBPP benchmark equity curve.

[MachineGhost]

MG: I am not clear how you scored them, but this one looks the best:
GSD Weight 60d, 10% Target GSD 60d, AbsMoMo  11/02/70  8.45%  -8.55% 

That ranked only 4th best.  The score is a measure of how much gain per unit of risk.  Score is more important than the absolute value of the CAR or MaxDD as they can be [de]levered during the Target stage.  Also, the 1% MG risk is not normalized to be equivalent to 10% of the others.

It would be interesting to see the equity curve & drawdown over the HBPP benchmark equity curve.

"Equal Weight" down near the middle-bottom at .80 score is the HBPP.

Here's the equity curves for the fourth and HBPP respectively:

[align=center][/align]

[align=center][/align]

Unfortunately, they don't come out right in logscale so they are linear.

[Stefan]

Nice  .  Looking at the charts, risk management made a big difference in 1981-82, 2000-01 and 2008.

[MachineGhost]

Nice  .  Looking at the charts, risk management made a big difference in 1981-82, 2000-01 and 2008.

The drawdown spike in late 2011 on the first chart bothers me.  Any idea what that could be from?

[Stefan]

Nice  .  Looking at the charts, risk management made a big difference in 1981-82, 2000-01 and 2008.

The drawdown spike in late 2011 on the first chart bothers me.  Any idea what that could be from?

From this - the waterfall drop in GLD:



I would look if a monthly rebalancing would have mitigated that.

[MachineGhost]

I have good news and bad news about the charts and stats I've recently posted.

The bad news is the simulated cumulative returns from the TNotes and TBonds were overstated.  I speculate that this was my fault from not realizing I should have added a 1 to each day's return before deriving the cumulative product in Excel.  I've coded it up so that R now calculates that automatically (which is how I found the discrepancy) so it won't happen again.  (To err is human, but to compute is divine!)

The good news is that the S&P 500 data I've been using was not total return as I originally thought.  I've compared several sources and no one is providing total return S&P 500 data (the CAR discrepancies are explained by different starting dates), so I've switched to using the Wilshire 5000 total return index and the increased gain from that offsets the newly decreased returns from the fixed income.  Overall PP portfolio risk has gone down too.

[melveyr]

I have good news and bad news about the charts and stats I've recently posted.

The bad news is the simulated cumulative returns from the TNotes and TBonds were overstated.  I speculate that this was my fault from not realizing I should have added a 1 to each day's return before deriving the cumulative product in Excel.  I've coded it up so that R now calculates that automatically (which is how I found the discrepancy) so it won't happen again.  (To err is human, but to compute is divine!)

The good news is that the S&P 500 data I've been using was not total return as I originally thought.  I've compared several sources and no one is providing total return S&P 500 data (the CAR discrepancies are explained by different starting dates), so I've switched to using the Wilshire 5000 total return index and the increased gain from that offsets the newly decreased returns from the fixed income.  Overall PP portfolio risk has gone down too.

I have reconstructed data using FRED interest rates in the past so I understand how much of a pain that process can be! You might consider using the Ibbotson Yearbook. They have a LTT index that targets a 20 year maturity, not perfect but the data is of good integrity.

[MachineGhost]

I have reconstructed data using FRED interest rates in the past so I understand how much of a pain that process can be! You might consider using the Ibbotson Yearbook. They have a LTT index that targets a 20 year maturity, not perfect but the data is of good integrity.

Two questions: Is it daily and is it free?

[melveyr]

I have reconstructed data using FRED interest rates in the past so I understand how much of a pain that process can be! You might consider using the Ibbotson Yearbook. They have a LTT index that targets a 20 year maturity, not perfect but the data is of good integrity.

Two questions: Is it daily and is it free?

Monthly and is commonly in libraries, so you gotta punch it into your spreadsheets by hand 

[MachineGhost]

Monthly and is commonly in libraries, so you gotta punch it into your spreadsheets by hand 

Hmm, tempting.  I'll check my local library and see if they have it.  I do have average yearly yield returns from Ibbotson back to 1880 that I got from here, probably you, clive or D1984.

EDIT: Looks like I'll have to make a trip to the Big City to access it.

[MachineGhost]

Using the new stock and bond data, MinCorr indicates the optimal portfolio since 1970 is:

Stocks: 66.0169%
Bonds: 26.89077%
Gold: 7.092243%

With TNotes:

Stocks: 23.6414%
Bonds: 9.745444%
Gold: 8.644297%
Cash: 57.96886%

With TBills:

Stocks: 6.084589%
Bonds: 3.001833%
Gold: 2.932052%
Cash: 87.98153%

[Stefan]

MG: There is no such thing as "optimal weights". And no such thing as historical correlations or historical volatility. These are variables oscillating in a wide range. So, looking at an average is like looking at a guy with his head in the oven and his feet on ice and saying that he should be OK on average  .
The best you can do with correlation and volatility is to use their momentum and  estimate their value at T+ k based on their measured values at T. And calculate weight(T+k) = f(StDev(T), Corr(T)) for each asset. Where k < 3 months. The smaller k, the better the estimate.
This is why the concept of "optimal weights" for a long period of time is a fallacy -and you should avoid this trap. Better go with PP equal weights, which is the agnostic way to deal with the unknown volatilities and correlations values at each moment(T).
Look at this post below and the curves I posted:

The difference between standard deviation and drawdown minimization with MinCorr is interesting.  The optimal weights is:

Stocks 13.98%
TBonds 7.27%
Gold 2.03%
TNotes 76.72%

...for 7.20% CAR and -5.10% MaxDD.  It can be levered by 68% to reach -10% MaxDD. Without the TNotes, it is:

Stocks: 60.86%
TBonds: 33.28%
Gold: 5.86%

...for 7.86% CAR and -30.56% MaxDD.  It can be delevered by 62% to reach -10% MaxDD. 

MG: I am not sure what you mean when you show "the optimal weights"?

There are no universal optimal weights, outside the rebalancing period. So, for instance, if your rebalancing periodicity is monthly, and you rebalance for risk parity and/or minimal correlation, you can calculate each month the optimal weights => for the next month. But, these optimal weights for month T will not be valid for month T+1. Because both correlation and volatilities, which are the input parameters in the rebalancing function, will change from the previous month...So, that "optimal weights" value set was confusing for me.

Here is the StDev for S&P 500 and its correlation to TBonds. You cannot talk about an optimal weight except for a period where these numbers change is relatively small, which is rarely > 1 month.

[MachineGhost]

That seems all good from a theoretical perspective where volatility = risk since it is so unstable, but I want to point out that the TNoteed weights are so "optimal" that no portfolio targeting or technical analysis further improves upon the risk-adjusted returns.  In fact, many worsen the outcome dramatically.  Was this an example of curve fitting?

Its true that MinCorr should be done in-line during rebalancings rather than ex post, but I can't do that automagically yet.

I do understand the basis of not using long-term averages due to short-term market changes, but I remain unconvinced it is bad practice when using something other than volatility for risk** or alternatively suggesting equal weight which is just as suspectible to the same swings.  Overweighting to gold as the equal weight PP does is not an agnostic bet.  Wouldn't equal risk be the true agnostic, ignoring the flunctuating correlations?

** What I mean by this is I'm concerned with the worst correlated multi-asset drawdown of the portfolio at a time, not whether or not the tracking error to shorter-term correlations and risk gets out of whack.  So what if it does?  It is mean-reverting back to the long-term averages.

[MachineGhost]

It seems that MinCorr has been bested already:

[align=center][/align]

Stefan, you should have told me about clustering as that is the rational I've been grasping for.   The disparity between weighting schemes overweighting low-volatility assets like bonds and equal weight overweighting riskier assets like gold was driving me nuts!  The subjectivity involved is one thing equal weight does not have.

So now that HB has been turned into an algorithm, he can finally rest in peace.

[Stefan]

Clustering is something I have not tested yet, I prefer to post only about algos I tested and run in real time.

P.S I think that, as this point in the discussion, most people following it are lost  ... But the principles of risk management I posted above should be universal enough and simple enough to be implemented via a variety of techniques and software, starting with very simple Excel spreadsheets - up to your own level of math skill and computer language literacy.

Those principles should be the main takeaways from my posts in this thread. Their implementation is dependent on your mileage.

[MachineGhost]

Clustering is something I have not tested yet, I prefer to post only about algos I tested and run in real time.

I am just getting into it, but my gut feel is that long-term clustering (i.e. back to 1970) will wind up almost like the PP.  So it would serve as the principal "top down" master diversifier and then you can use the other weighting schemes discussed in this thread within the major cluster groups.  If clusters truly represent distinct economic environments, then there should be no instability.  For instance, since 1996:

[align=center][/align]

Here we see that the clustering is four major groups, but that ST bonds are within the fixed income cluster.  That's an interesting find and may point the way that TBills/TNotes and actual U.S. cash currency do not act the same, but I guess that is in keeping with the PP's barbell concept.  It also looks like foreign stocks are distinct enough not to be lumped into a domestic equity cluster (which also clusters REITs and precious metal stocks which I would expect).  The stock-bond funds are lumped in with domestic equity as that is literally true in reality, for they are 90% exposed to stock risk.

And then normalizing the clusters as well as the assets within the clusters by risk we see how unbalanced equal weight truly is:

[align=center][/align]

Very gratifying even if it is sort of obvious the AWP/PP inspired the implementing of clustering to backfit the theoretical "economic environment" principles into practical quantification.  But I'm ecstatic! 

And since the peak in 2007 we can see that the weights are more or less stable:

[align=center][/align]
[align=center][/align]

P.S I think that, as this point in the discussion, most people following it are lost  ... But the principles of risk management I posted above should be universal enough and simple enough to be implemented via a variety of techniques and software, starting with very simple Excel spreadsheets - up to your own level of math skill and computer language literacy.

Those principles should be the main takeaways from my posts in this thread. Their implementation is dependent on your mileage.

Good points.  So far in my backtesting, the technical analysis stage has had the most impact on upping portfolio return without adding to the risk.  That's unfortunate as technical analysis is problematic to implement on the traditional PP diversified for maximum safety.  And we still haven't settled the issue of the impact of transaction costs and taxes which need to be factored in (perhaps that Composite Standardized Score would be helpful here).  I'm also wont to point out that before 1996 which could be considered out-of-sample, the "risk moderation" principles don't seem to have held up very well, and warrants closer study.

EDIT: The grouping into four major classes is an artifact of the stock-bond funds being included.  With them removed, there are only three major groups, i.e. equity, fixed income, real with no split between domestic/foreign equity.  Also, adding TBills:

[align=center][/align]

[align=center][/align]

So TBills appear to be distinct enough from longer duration Treasuries.  It may be a mistake to conflate the two in actual implementation.  But I would not draw any conclusions based on such limited data.

[MachineGhost]

I just a did an iteration over all possible 1,512 portfolio compositions for the vanilla PP assets (w/TNotes) since 11/1970 and what winds up being the most reward/risk efficient is:

Rebalancing: 6-month
Weights: Equal
[De]Levering: 10% GSD Target (3-month)
TA: Trend Following (200-days)
Return/Risk: 8.34% CAR,. -11.56% MaxDD

I included .26% for market impact and allowed up to .5% on top of that for commission costs per trade.  That's the best I can do right now to simulate transaction costs.  (Taxes would amount to around .51% a year but I'm not able to deduct that just once and it would penalize non-taxable accounts anyhow.)

The 1,512 portfolios do not include recently mentoned techniques such as clustering, MinCorr or MinVariance, but as they are only "risk moderation", its not too much to worry about.  The best "risk moderation" techniques will reduce the risk of the vanilla PP by almost half at moderate cost of some return.  The "risk reduction" techniques will then bring back up the return.

For those adverse to TA for whatever reason, the best non-TA portfolio (w/TNotes) was:

Rebalancing: 12-month
Weights: MaxDD
[De]Levering: 10% GSD Target (3-month)
Return/Risk: 7.87% CAR,. -12.09% MaxDD

The traditional PP (w/TBills):

Rebalancing: 12-month
Weights: Equal
Return/Risk: 8.10% CAR,. -22.58% MaxDD

[iwealth]

How did your system react to the 200-day trend following? Sell when a trendline was breached, buy on a break-out above? If you don't mind me asking, that is..

[MachineGhost]

How did your system react to the 200-day trend following? Sell when a trendline was breached, buy on a break-out above? If you don't mind me asking, that is..

Just a simple price crossover of the 200 MA.  It'll probably be better to use a monthly closing price and a 10-month MA to avoid whipsawing.

MA crossovers aren't supposed to have worked since the 70's when rising inflation produced long-term extended trends in commodities, yet its hard to argue with the theoretical performance ex post facto.  In reality, I do believe the higher transanctons costs and taxes would have made them unprofitable vs buy and hold to actually trade, but the situation may be different now.

[MachineGhost]

I almost hate that I found these ETF's, but if Goldman Sachs is going to be useful for something besides ripping off muppet faces, then it might as well be for this:

http://www.alpsgsetfs.com/momentum-builder.php

[melveyr]

I almost hate that I found these ETF's, but if Goldman Sachs is going to be useful for something besides ripping off muppet faces, then it might as well be for this:

http://www.alpsgsetfs.com/momentum-builder.php

What makes you think that these ETFs aren't designed for ripping off muppet faces?

Somewhere along the line you got the perception that muppets like easy, transparent, and low cost solutions. I don't think that has ever been true. Muppets get ripped off when they think they are smarter than they really are and buy overly engineered solutions without understanding the hidden risks they are taking.

[MachineGhost]

What makes you think that these ETFs aren't designed for ripping off muppet faces?

Somewhere along the line you got the perception that muppets like easy, transparent, and low cost solutions. I don't think that has ever been true. Muppets get ripped off when they think they are smarter than they really are and buy overly engineered solutions without understanding the hidden risks they are taking.

That's quite possible, but nothing in the ETF's seem any different than what has been discussed earlier in this thread.  I've got a hard time believing that the ETF's are suddenly attractive to the muppets and that GS wants to rip their faces off using it.  Aren't there much easier ways for GS to skin the muppets?

But, suddenly the investing landscape has become more competitive, especially for the staid ol' PP.  It's going to have to run for its money to keep up if those ETF's are sign of things to come.  We can only win in the investing game long-term by at least being average and if the average turns into sophisticated risk control mechanisms without lifting a finger for 1% fees or so that any muppet can tap into, PPers will risk become sub-muppets!

[melveyr]

But, suddenly the investing landscape has become more competitive, especially for the staid ol' PP.  It's going to have to run for its money to keep up if those ETF's are sign of things to come.  We can only win in the investing game long-term by at least being average and if the average turns into sophisticated risk control mechanisms without lifting a finger for 1% fees or so that any muppet can tap into, PPers will risk become sub-muppets!

I guess I'm surprised how excited you are about these Goldman Sachs ETFs. Ironically, you might end up being the biggest muppet of all 

[MachineGhost]

I guess I'm surprised how excited you are about these Goldman Sachs ETFs. Ironically, you might end up being the biggest muppet of all 

I'm not unaware of the irony.  But I wouldn't call it excitement, more like dread.  At the same time anything that could potentially ease my workload is very welcome.  So lets call it cautious optimiim.  But as with PERM, proof will first be needed.