Thread: A Study for Griffin

Desperado,

No matter what they say, keep it up.

Analysis is a good thing if it makes you reassess your investment strategy.

Even I am not happy with my investment strategy for this year, but I have been formulating a new one for next year.

Pointman72

Desperado,

You did not quite capture the essence of what I was getting at in your analysis. I will get to work on a spreadsheat that shows you what I mean. But I have to put that off until later. I will get back to you tonight or tomorrow on this.

Your patience is appreciated

Originally Posted by pointman72

Even I am not happy with my investment strategy for this year, but I have been formulating a new one for next year.

Yep, you were slackin' with that 24.5%. 30% would have been better.

Originally Posted by pointman72

...
Even I am not happy with my investment strategy for this year, but I have been formulating a new one for next year.

Pointman72

Hey Pointman:

If you're not going to be using your strategy, would you mind if I borrow it for next year?

Originally Posted by Desperado

Okay, here's the numbers you asked for Griffin: 1996-2005
Fund Annualized Return
G 5.49%
F 6.18%
C 9.01%
S 9.74%
I 5.67%
20% Each rebalanced once a year 7.77%
All Stock (1/3 C, 1/3 S, 1/3 I) 8.21%
Picking the best fund Jan 1 and holding it until Dec 31 21.99%
Picking the worst fund Jan 1 and holding it until Dec 31 -4.54%
Underperforming the best fund each year by 13% 7.93%

Looks like a pretty good strategy, huh? You can miss the top fund by 13% and still outperform a simple buy and hold 20% each! 70% of the people I studied beat that this year. That looks pretty good. But let's turn it around a bit and examine it a little closer. What if we measured outperforming the worst fund, rather than underperforming the best fund? How much do you have to beat the worst performer by to have a great return? It turns out to beat a simple 20% each buy and hold strategy you have to beat the worst performer by 13% also. The worst performer this year was the G fund at 4.48%. Add 13% (17.48%) and it turns out only 10% of you beat that. The reason for the quirky data? The huge bull market in the late 90s and the big bear in 00-02.

Bottom line, don't compare your returns to the top performing fund each year. It is a lousy benchmark. I submit that the proper benchmark for most of you is probably 20% each. If you are consistently beating that (year after year), perhaps you have some talent and it is worth your time to churn your TSP account. If not, well, maybe your time would be better spent doing something else.

shewww...finally (20/20/20/20)/20). I would add the L2040 would also be a reasonable benchmark as it represents the mainstream touted diversified, buy n hold strategy. Debating the better of these two would be of more interest/value.

Fundsurfer,

Problem is I could have made an additional 17% if I would have moved to the G fund, instead of staying invested when I saw the market turning. Got entirely to focused on trying to out perform the I fund instead of maximizing my return. By the way, started posting my moves in March and did make 2 percent in Jan/Feb.

James,

I'll share my new strategy with everyone when I finish sorting it out, but simply it will be based on a buy and hold of the I fund except for movements to the C/S fund when the dollar is dropping. Only reluctance is that Asia and Europe have had a hard time synchronizating their efforts as of late.

Desperado,

You did what I asked, just not in the format I was expecting. Not your fault, I'm one of those guys that needs to see the raw data to draw my conclusions. I compared my calculations to yours, and I believe we came up with the same data. Here is mine, you can confirm.

Observation 1) The best performing fund over the 10 year study period was the S-fund yielding 253%. With a 2.5% inflation rate, the original investment amount would have to increase by 128% just to break even (inflation adjusted dollars). Therefore, with a 100% buy and hold strategy in the S-fund, you would effectively doubled your money in a 10 year period.

Opinion - A 10 year inflation adjusted doubling rate for a young investor is mediocre at best and this is the best buy and hold strategy you could have had with the benefit of hindsight. My feeling is that a young investor needs to be doubling their money every 6-7 years, which is a reasonable expectation for a decent aggressive buy and hold stock portfolio, adjusted annually.

Observation 2 (same as yours but focusing on the best buy and hold scenario) A timer who lagged the best performing fund each year by 12% would have achieved about the same buy and hold yield. That same timer could have achieved that goal by beating the worst performing funds each year by 15%. Basically the same quandry you presented for 2006. The majority of us are within 12% of the I fund, but very few of us are 15% ahead of the G fund.

Opinion - What is the worst performing fund data telling us? It says we must have outstanding years during the bull years to compensate for the enormous losses we are going to take in the bear years. Subsequently, the best performing data tells us we can get away with lagging the market in the bull years, as long as we do not get crushed in the bear years.

Your conclusion is that the bull benchmark is the wrong benchmark and that an average benchmark is more appropriate.

With that in mind, go back and look at Tom's returns. He lags the market (only slightly much of the time), as he is now during a bull year, but he does not get crushed in the bear years. My observations of my own behavior is that I tend to slip behind the stock funds during the big bull weeks, and play catch up during the bear weeks. Based on the behavior patterns I am observing, the best performing data results seem to describe the behavior we are observing in the field.

Also, I would like to draw to your attention, the fact that the same timer could have quadrulped their return if they lagged the best performing fund by 3% or less. They could have tripled their return if they were within 8%. I don't know the retirement horizon for all our members, but I have a general idea about enough of them that I am willing to propose that the majority of our young aggressive investors are on target for the later goal (with the right support network and some more experience). Given that most people here are still in the learning stages, I would call the site a preliminary success, but only time will tell. You may not agree with what I have to say, but I think you can see where the motivation comes from.

The last thing I want to mention, is that the 20% in each allocation would have only slightly better then doubled an investor's money WITHOUT adjusting for inflation over the ten year period - that IS abyssmal.

Perhaps a consensus is more appropriate and we should be looking at the performance of the timers in ranges from .... let's say the average +4% to -the best - 10%. Call those folks "success", those above it "outstanding". A slightly lower range of Average to the Best - 14% "no effect" and those below, "negative impact". How does this sound?

Originally Posted by pointman72

Problem is I could have made an additional 17% if I would have moved to the G fund, instead of staying invested when I saw the market turning.

I was being sarcastic. You did very well and we can always look back and see how we can improve. I look at the maximum possible return for a year (if you picked the right fund to be in every day) and think we have a lot of room to improve.

Originally Posted by Griffin

Opinion - A 10 year inflation adjusted doubling rate for a young investor is mediocre at best and this is the best buy and hold strategy you could have had with the benefit of hindsight. My feeling is that a young investor needs to be doubling their money every 6-7 years, which is a reasonable expectation for a decent aggressive buy and hold stock portfolio, adjusted annually.

Opinion - What is the worst performing fund data telling us? It says we must have outstanding years during the bull years to compensate for the enormous losses we are going to take in the bear years. Subsequently, the best performing data tells us we can get away with lagging the market in the bull years, as long as we do not get crushed in the bear years.

Your conclusion is that the bull benchmark is the wrong benchmark and that an average benchmark is more appropriate.

Also, I would like to draw to your attention, the fact that the same timer could have quadrulped their return if they lagged the best performing fund by 3% or less. They could have tripled their return if they were within 8%. I don't know the retirement horizon for all our members, but I have a general idea about enough of them that I am willing to propose that the majority of our young aggressive investors are on target for the later goal (with the right support network and some more experience). Given that most people here are still in the learning stages, I would call the site a preliminary success, but only time will tell. You may not agree with what I have to say, but I think you can see where the motivation comes from.

The last thing I want to mention, is that the 20% in each allocation would have only slightly better then doubled an investor's money WITHOUT adjusting for inflation over the ten year period - that IS abyssmal.

You draw different conclusions from similar data. That is your right. I think this side study is seriously skewed by the atypical data from 96-03. I bet if we widened it out to include the entire TSP period it would be more useful. I think it is a serious investment mistake to benchmark your return to the highest asset class return minus a certain percentage. What if you decide that if you can just stay within 12% of the top performing fund you'll meet your investment goals, and then for the next ten years NOTHING returns more than 15% a year, you could have poor investment returns and yet still beat your benchmark.

Now, I want to argue with your opinions. Doubling an investor's money in real terms every ten years is an excellent return, not an abysmal one. In fact, it is a better return than the long term returns of the S&P 500. I disagree that an investor needs to double his money (in real terms) every 6-7 years to reach his goals. Also, I DON'T feel that that it is a reasonable expectation for a decent aggressive buy and hold portfolio. I think a reasonable expectation for an aggressive buy and hold portfolio is 9-10% (6-7% real). Doubling your money every 6-7 years is the equivalent of a 10-12% annualized return, without counting inflation. Counting inflation, we're talking 13-16%. That is a fantastic return, and much better than I am planning on with my buy and hold investment portfolio. Let me tell you just how fantastic a 12% real return is, if an investor can pull it off.

An investor who saves a mere 10% of his income each year can retire after only 29-30 years of work and easily maintain his pre-retirement lifestyle (4% withdrawal rate) without any assistance from pensions or social security. Hell, if you can pull off 12% real returns consistently, you don't even need to limit yourself to a safe 4% withdrawal rate. Why not have a 8% withdrawal rate? Then you could retire after a 24-25 years. Since most investors will have at least social security, they could work even less years. Certainly this return is much better than is required for the typical investor. Perhaps those who start saving late or don't save much of their income need a 12% real return, but most of us shouldn't. I expect to fulfill my investment goals with a 24 year career, a 25% savings rate, and 5% real returns, without social security.

12% real return is extremely optimistic, especially for a buy and hold investor. The historical S&P 500 real return is 6.5-7%.

Real returns for the TSP funds over the last 10 years have been:
G Fund: 2.7% (money doubles every 27 years)
F Fund: 3.4% (money doubles every 21 years)
C Fund: 6.2% (money doubles every 12 years)
S Fund: 6.9% (money doubles every 10 years)
I Fund: 2.9% (money doubles every 25 years)

How do you figure you can pull off 12%? I think that is exceedingly optimistic and is an idea that was heard a lot during the roaring 90s. If your investment plan depends on 12% real returns I submit that you are likely destined to fail, no matter how good your asset allocation or how good you are at market-timing.

As far as benchmarks go, I have been thinking that the best benchmark for you market timers would be to take the average return of the 5 funds on a daily basis, and combine them to make up the entire year's return. This would be the equivalent of performing against 1000s of monkeys who traded their portfolio every day.

The last idea I would like to dispute is this idea that many of the market-timers are just learning. So are some of the buy and holders. How long of a learning curve can one tolerate? If half of your investing lifetime is the learning curve, how much better do you have to do in the last half to make up for your underperformance in the first half when compared to a simple buy and hold allocation that you could have developed during the first week of your investing lifetime? How long does it take to be a proficient market timer? Why do some people still not seem to be able to do it despite the fact that they study the markets every day for decades and even design websites to talk to others about it? Because it's hard? Maybe. Because it's impossible and those who seem to have success doing it are actually just lucky? Seems like a more likely explanation to me. But give it time. We'll see in a few years if all those people learning are getting better.

Desperado,

I am not trying to be rude, but you are grasping at straws. I say that because you devoted the vast majority of your response, to explaining why my approach was not necessary, rather then not possible.

First I want to address the part about benchmarking. Look at a one decade chart of any decade (not a YTD chart spanning from some time period to the present), I mean a start and stop date 1 decade apart, with the Y axis expressed as a percentage. If you do that going back through the last century, you will very quickly realize, that the volatility of the 90’s was not that bizarre, but as you read through this post, you should realize that ultimately, it is irrelevant.

The way you express the benchmarking, you make it sound as if I set my benchmark at some relative position, I am going to actively make that my target. If my benchmark is the best fund – 5% and I am at the best fund – 1%, I am NOT going to go out and actively lose 4% to meet my benchmark. You suggested a set buy and hold strategy. The TSPtalk returns calculator spreadsheet is set up so that you can see how your return stacks up against a number of set of buy and hold strategies (one of which is the one you initially mentioned, 20% in each). The benchmark itself is not a goal or a guideline, it is simply a measure by which we can show that a reasonable possible return is achievable.

If I said our goal was to achieve the best return produced by the best performing fund each day, I would be setting a goal of approximately 200% a year (because that is what is theoretically achievable). However, that is insane, no one can get it perfect everyday. So far this year, I have made 50%, but I have lost 30% for a net gain of about 20%. That’s not luck, I stack the deck in my favor through systematic analysis, research and a knowledge of events and factors that flow steadily from the members of this board. Regardless of how volatile or static a year may seem on graph paper, the theoretical ability exists to make or loose 150-250% a year through daily volatility. All we have to do is to consistently stack the deck just enough (like a casino does), that we consistently skew the net returns in our favor such that we move the monkey dart median about a third of one standard deviation in our favor. The reality of the good timer, is that the annual returns of any index fund does not need to go up or down at all, ever, for us to achieve a 12% return, all it has to do is be volatile, and have measurable, semi-predictable behavior.

We all acknowledge that there is no set specific strategy that works under every condition. By timing we can be flexible and adaptive. The only purpose of the benchmark is to provide some generalized correlation that is representative of the fact that we should be able to book more gains in a bull year then in a bear year.

If you read through my account talk thread and the bull pen threads, I have mentioned on several occasions that when the market is moving up quickly, a buy and hold strategy is ideal. Conversly, if the market is in a consolidation/pullback, a capital preservation mode is best. The situation that you describe where the market never sees more then 15% for ten consecutive years (although that has never actually happened – maybe monkey’s will fly out my butt) suggests a prolonged sideways market ranging in an established channel. This, my friend, is timer heaven…..the great opportunity to buy low, sell high, go into capital preservation mode and repeat the process over and over and over, with virtually no lose to inflation.

Tom often brings up casino analogies in his market comments. This is the reason why. The last time I checked, RevShark had stock trading accounts that were up over 30% YTD.

You strike me as an intelligent person who has a very good grasp on numbers, odds and percentages. There is no doubt in my mind that you could be a successful timer.

Well said Griffin!