My trend following strategy and experience

[willthrill81]

I think whole thread is a collective fail by the Boglehead community. Why the deference and applause for a market timing scheme? Why aren't forum veterans pointing out that buying and selling of an entire portfolio based on some economic indicators is ill-considered?

Mostly because (1) it's not our money, (2) he saves a lot and can afford to under-perform, and (3) he's not leveraged. It is not ideal, but the OP has already stated he's not interested in optimization, and if this is what he wants to do, it's his choice. I don't agree with hit, and based on the replies, neither does most everyone else. But the worst case scenario here seems to be opportunity costs and underperformance, as opposed to tanking his entire retirement account with leverage and derivatives.

Again - I don't agree with this strategy, and think it is likely to hurt him in the long-term. But I also don't see the point in moral condemnation. If we can't talk him out of it, we may as well wait to see how it turns out.

Moral condemnation seems like an extreme characterization. It's more about sending the usual message that market timing, and especially extreme shifts, is likely to be unsuccessful.

Not to mention the incumbent was released from his equities 200 points lower.

I'm not at all concerned about a single whipsaw or even a series of them. I've studied the history of the markets and strategies like mine to know that that's inevitable, just as deep drawdowns are an inevitability of buy-and-hold (and might still happen to me to). That's a price that I'm willing to pay, just as many here are willing to forego the likely outperform of stocks for the stability of bonds.

Further, you could apply the same argument to those with value tilts and U.S.-only portfolios. Both are sure to underperform at some point in the future.

[willthrill81]

I'm not quite sure how your system is better than just having some money in fixed income all the time. When you are 100% out of stocks, you are 100% in fixed income, so how would this system have protected you in 1977-1981?

I'd have to go back and look at the data for that period, but I would have been in short-term bonds, which fared much better than longer-duration bonds during that period, or stocks, which mostly treaded water.

[willthrill81]

Jack Bogle famously said he didn't know anybody could successfully time the Market and didn't know anybody who knew anybody who could successfully time the Market.

Well he knew Paul Merriman, and Merriman has used a trend following strategy with half of his portfolio since the early 1980s. He reports achieving very similar absolute returns to buy-and-hold since then but with lower drawdowns.

[willthrill81]

I think whole thread is a collective fail by the Boglehead community. Why the deference and applause for a market timing scheme? Why aren't forum veterans pointing out that buying and selling of an entire portfolio based on some economic indicators is ill-considered?

Mostly because (1) it's not our money, (2) he saves a lot and can afford to under-perform, and (3) he's not leveraged. It is not ideal, but the OP has already stated he's not interested in optimization, and if this is what he wants to do, it's his choice. I don't agree with hit, and based on the replies, neither does most everyone else. But the worst case scenario here seems to be opportunity costs and underperformance, as opposed to tanking his entire retirement account with leverage and derivatives.

Again - I don't agree with this strategy, and think it is likely to hurt him in the long-term. But I also don't see the point in moral condemnation. If we can't talk him out of it, we may as well wait to see how it turns out.

Moral condemnation seems like an extreme characterization. It's more about sending the usual message that market timing, and especially extreme shifts, is likely to be unsuccessful.

You won't find me recommending this or any other trend following strategy to anyone here ever. When I do offer recommendations to others, it's always based on a buy-and-hold strategy.

I've done my best to carefully lay out the reasons for my chosen strategy with all of the caveats pertaining thereto, including my acknowledgement that I may completely fail in my desired objectives. Doing so has laid me open to criticism, and many others in this thread have already taken advantage of that, which is perfectly fine and an expected consequence on this type of forum. At the worst, my example will serve as a warning to others of the pitfalls of this strategy.

[28fe6]

Making a bunch of little bets when the odds are in your favor is smart gambling; betting your entire bankroll is not

This is why I developed my "never sell" policy (viewtopic.php?t=271567). I typically know that trying to time the market is folly, but my never-sell policy allows me to make any bet I think is the right bet, or use any system that I think is the right system, so long as I do it with new money. I can still time the bejeezus out of the market with my contributions, but I do not allow myself to sell securities in order to layer "change my mind" about what I bought.

Since December I have been buying zero bonds and scraping up every penny out of the couch for stocks in order to buy the dip. There was a period a year or two ago when I temporarily got skittish and bought only bonds because I thought stocks were over-valued. Both were "small bets" compared to the destruction I could have caused by selling and re-allocating.

I am considering incorporating the OP's trend-following method with my contrbutions, while still sticking with my overall "no-sell" policy. However it seems for contributions, the reaction should be reversed (buy more stocks when indicators are bearish, buy a more bond-heavy allocation when indicators are bullish.

[willthrill81]

100% swings seem really extreme to me.

I was 100% stocks before I adopted this strategy, so swinging from something like 25/75 to 75/25 isn't appealing to me.

I'm convinced enough in the forward success of my strategy to go all-in. When I get closer to retirement, I might back off of that though.

[3504PIR]

I applaud your efforts and willingness to be a potential punching bag here. I also wish you luck in this effort and I’m curious to see how you do. What assets are you buying and selling? Is your intent to capture a traditional trend following strategy or a momentum strategy? I’m not an expert, but what you’ve written seems to be in the later category. The previous reading I’ve done on trend following is that the strategy is wide open to a variety to asset classes to include commodities. Feel free to correct me on this.

Good luck again and I look forward to following.

[nps]

100% swings seem really extreme to me.

I was 100% stocks before I adopted this strategy, so swinging from something like 25/75 to 75/25 isn't appealing to me.

I think your point is that when you've settled on a new strategy you'd prefer to be all in.

If I understand your intention for trend following correctly, it's designed to minimize the likelihood that you'll decide to bail out in a severe downturn. That's also the role fixed income can play in a buy-and-hold asset allocation.

You've argued previously that AAs around 80/20 provide no behavioral benefit for avoiding panic selling. That view may also shed some light on why you favor your current strategy. Have you simply concluded that the risk of tracking error regret under your current strategy is less significant than the risk of panic selling under a fixed buy-and-hold AA?

[BlackStrat]

I also enjoy your posts and would wish you the best but that would mean my 3-fund approach was inferior!

(But I certainly don't wish you failure) Good luck!

[thx1138]

Thanks for sharing Will.

As to the "is this Bogleheads" or not recall that the real fundamental basis to Bogle is the "costs matter hypothesis". As implemented the costs for this strategy are no different than "buy and hold". Bogle himself advocated many AA changes in various interviews that are easily interpreted as "market timing" by the more zealous on this forum. What matters is not paying excessive fees or excessively trading. This strategy clearly passes that test.

Note that "buy and hold" is not in fact "buy and hold" as typically implemented by Bogleheads. There is typically rebalancing and some sort of AA adjustment over time. A key point the article this strategy originates from makes is that this strategy over time is roughly an AA of 85/15. The only difference is that this AA is divided up in time. You spend 85% of your time in 100% equity and 15% of your time in 100% FI. This is essentially an extreme form of rebalancing. All rebalancing strategies involve either time thresholds or valuation thresholds to govern trades - this strategy is different in magnitude.

Again compared to typical rebalancing strategies this is still on the "simple" side of things. The monthly checks are no more onerous than checking typical rebalancing thresholds (like 5/25). The number of trades executed is no more - and likely actually less - than typical rebalancing strategies.

I wonder about the "behavioral" aspect of this though. Not selling in a downturn may be hard but it seems "getting back in" when your strategy tells you to may be just as challenging as a buy and holder needing to not sell and rebalance into equities. But this is all a very personal kind of thing so it probably fits into the "whatever makes you sleep better at night" bin. Unfortunately though you won't actually know the reality of it until tested - which is of course just the same as a buy and holder.

Lastly for those obsessing over the UER moving average crossing. Fundamentally the rounding used is irrelevant as it just results in a small fixed offset in the comparison. The particular period of the average and the particular UER metric used are largely arbitrary. The particular rounding used is similarly arbitrary. The discomfort seems to be better placed not on the rounding but rather that this strategy is binary - all or nothing - based on a threshold. Regardless of the metrics and math chosen binary strategies always have this discomforting behavior. That is a cost of keeping them relatively simple and again the reality is that in the end whether you choose one slightly different metric or slightly different threshold the over all results are roughly similar. That is in fact a robustness test for any strategy - small changes in the strategy result in similar outcomes. Again this particular strategy as illustrated in the original article is robust against such variations. Being a binary time based strategy though the short term variations can be quite significant but of course anything heavy in equities is going to have large variations on short time scales.

Not my particular cup of tea to follow this strategy but I don't see it as a particularly large philosophical deviation from most of the more typical BH approaches on the forum. The "tracking error" such as it is will likely be no more than the difference between those who invest in ex-US equities and those that don't.

Keep us updated Will!

[sixtyforty]

I personally think the trend following strategy has merit, as I follow a very simple one myself. I think you put it best... willing accept a bit of under performance due to whipsaws that may occur, to avoid the big draw-downs. The simpler the strategy the higher the probability it will work in the future. It's why BH works because it's simplest of ALL. No parms to modify. You buy and hold forever... IMO, the more parms or decision points a strategy uses the higher the probability of failure.

While a lot of people are focused on discussing the trend following, it's actually the relative strength part of the strategy (IMHO) would cause me concern. I've never found one that works in the future. They can back-test well but the parms used is rarely the correct parms for the future. For example, I think you are using 7 months, which may or may not be the correct parm to measure RS. When it's not, you might be tempted to change it and then you are caught driving forward by looking in the rear view mirror.

Here is an interesting live study... The link below shows an actual relative strength strategy that has been ongoing for a number of years. The strategy believes that there is always one best investment for a time period, which is true, and uses RS. However, determining that one investment is not easy. Don't be fooled by the chart. Take a look from 2009 forward where the strategy has basically gone nowhere for 10+ years. Investing back then you would had to endure volatility and not have much to show for it. Not to mention, he's been tweaking the strategy along the way to try and improve it's under-performance.

http://www.decisionmoose.com/Moosistory.html

[BlueEars]

...
Lastly for those obsessing over the UER moving average crossing. Fundamentally the rounding used is irrelevant as it just results in a small fixed offset in the comparison. The particular period of the average and the particular UER metric used are largely arbitrary. The particular rounding used is similarly arbitrary. The discomfort seems to be better placed not on the rounding but rather that this strategy is binary - all or nothing - based on a threshold. Regardless of the metrics and math chosen binary strategies always have this discomforting behavior. That is a cost of keeping them relatively simple and again the reality is that in the end whether you choose one slightly different metric or slightly different threshold the over all results are roughly similar. That is in fact a robustness test for any strategy - small changes in the strategy result in similar outcomes. Again this particular strategy as illustrated in the original article is robust against such variations. Being a binary time based strategy though the short term variations can be quite significant but of course anything heavy in equities is going to have large variations on short time scales.
...

I think you're referring to my post. I agree that much of my discomfort is in the consequences of the binary nature of the decision. I personally am comfortable with an all out approach but actually doing it would be uncomfortable for me. I guess that is a rational decision that requires emotional strength to implement.

So I'm just looking at the UER usage to try to see if it is robust. Unfortunately I don't think it is all that robust because of (1) BLS precision (2) rounding error, and (3) possible government revisions. Probably the government revisions is the most troubling. Suppose you went all out and then found out a few months down the road that they revised one of those numbers for the 12 month moving average calculation and you should have stayed in the market. Not so bad if you did not get whipsawed but costly if you did. So I'd use the data but put a somewhat higher threshold on this before pulling the trigger. Now with the market pricing data for moving average calculations, I'd be comfortable with that.

[H-Town]

Jack Bogle famously said he didn't know anybody could successfully time the Market and didn't know anybody who knew anybody who could successfully time the Market.

Well he knew Paul Merriman, and Merriman has used a trend following strategy with half of his portfolio since the early 1980s. He reports achieving very similar absolute returns to buy-and-hold since then but with lower drawdowns.

Are you really scared of big draw-downs? Those draw-downs are necessary for bigger run-ups.

If your goal is to avoid draw-downs, holding 100% Wellesley would be a smarter way to go.

Don't buy and hold 100% stock, buy and hold a well balanced portfolio.

[Forester]

Between 20% to 50% is a smart trend following allocation.

Let's say you're 40/40/20 B&H/trend/bonds. You would be 60% in risk-off assets when/if the market rolls over. Surely that's good enough? And one wouldn't be moving 100% of their portfolio all in one go. I think a year of a few whipsaws would severely test one's ability to stay the course.

Also - you probably get some kind of synergy from mixing buy & hold with trend. I've thought about this a lot, and I think another decade like the 2000s, of two V-shape 50% drops, is unlikely. It could happen again but not for half a century.

[willthrill81]

I applaud your efforts and willingness to be a potential punching bag here. I also wish you luck in this effort and I’m curious to see how you do. What assets are you buying and selling? Is your intent to capture a traditional trend following strategy or a momentum strategy? I’m not an expert, but what you’ve written seems to be in the later category. The previous reading I’ve done on trend following is that the strategy is wide open to a variety to asset classes to include commodities. Feel free to correct me on this.

Good luck again and I look forward to following.

The details on my strategy are in the OP.

And yes, trend following has been demonstrated to have worked successfully in an array of markets, not just stocks. The Philosophical Economics link I posted discusses this.

[willthrill81]

If I understand your intention for trend following correctly, it's designed to minimize the likelihood that you'll decide to bail out in a severe downturn. That's also the role fixed income can play in a buy-and-hold asset allocation.

That's true, but fixed income over the long-term is basically just ballast. Over the last decade, you would have done far better to have held none. So I would be concerned about dealing with the regret of holding too much stock when they're headed down or not holding enough when they're headed upward.

You've argued previously that AAs around 80/20 provide no behavioral benefit for avoiding panic selling. That view may also shed some light on why you favor your current strategy.

To be clear, I've never said that they hold "no" behavioral benefit, only that most investors are unlikely to hold the course in the face of a 40% decline in their portfolio but panic at a 50% decline.

Have you simply concluded that the risk of tracking error regret under your current strategy is less significant than the risk of panic selling under a fixed buy-and-hold AA?

More or less, yes.

[willthrill81]

I also enjoy your posts and would wish you the best but that would mean my 3-fund approach was inferior!

(But I certainly don't wish you failure) Good luck!

Well, we could both be very happy. Over the long-term, I do not expect to have better absolute returns than a buy-and-hold strategy, just a smoother ride.

[willthrill81]

I think a year of a few whipsaws would severely test one's ability to stay the course.

Similarly, 2008 was a year that severely tested many people's ability to stay the course, though virtually every trend following system did well, with some even posting gains for the year.

Pick your poison: occasional deep drawdowns or more frequent whipsaws.

[sergeant]

and he is not bragging about how great he is doing with it.

That's a good point, and a prime difference between this thread and other market-timing threads we've seen in the past.

Market-timer's thread too was like this one, and it was better received than most for the same reasons.

It would be pretty hard to be bragging at this point. I would estimate he is down 5%+ in three weeks against his benchmark.

[Forester]

I have the seven month moving average and all of the available indexes already set up in Portfolio Visualizer. Examining the UER moving average can be done in about two minutes using the BLS data.

Hello, which link/feature do you use for this on PV?

Do you simply check the timing model page https://www.portfoliovisualizer.com/tes ... bol1=VFINX on the first trading day of the month? Thanks

[willthrill81]

I have the seven month moving average and all of the available indexes already set up in Portfolio Visualizer. Examining the UER moving average can be done in about two minutes using the BLS data.

Hello, which link/feature do you use for this on PV?

Do you simply check the timing model page https://www.portfoliovisualizer.com/tes ... bol1=VFINX on the first trading day of the month? Thanks

The timing model I use is "relative strength."

[hushpuppy]

willthrill81, I too can't stand the wild B&H drawdowns and am thrilled to see how your market timing effort pans out - thanks for starting this thread! Hoping you'll keep posting the entries/exits dictated by your strategy and its relative performance from time to time so that the rest of us can keep track of it.

You're welcome! If nothing else, this thread keeps the forum from being an echo chamber.

I'll post every time I move in/out of the market.

willthrill81, I too am concerned about large drawdowns and appreciate your willingness to put yourself in the arena with this thread. Some apparently think they should give an early thumbs down. However, I hear and agree with your sentiment:

keeps the forum from being an echo chamber

Not promoting an echo chamber environment is in itself valuable. Many people here appreciate hearing different points of view.

Even in elementary school, I use to detest repetition.

Thanks for an interesting thread and your patience with your responses. I am older and considerably crankier, than you are. Perhaps, I can learn from your example. And perhaps, some will realize that criticism is not as valuable a commodity, as they first imagined.

hushpuppy

[willthrill81]

willthrill81, I too can't stand the wild B&H drawdowns and am thrilled to see how your market timing effort pans out - thanks for starting this thread! Hoping you'll keep posting the entries/exits dictated by your strategy and its relative performance from time to time so that the rest of us can keep track of it.

You're welcome! If nothing else, this thread keeps the forum from being an echo chamber.

I'll post every time I move in/out of the market.

willthrill81, I too am concerned about large drawdowns and appreciate your willingness to put yourself in the arena with this thread. Some apparently think they should give an early thumbs down. However, I hear and agree with your sentiment:

keeps the forum from being an echo chamber

Not promoting an echo chamber environment is in itself valuable. Many people here appreciate hearing different points of view.

Even in elementary school, I use to detest repetition.

Thanks for an interesting thread and your patience with your responses. I am older and considerably crankier, than you are. Perhaps, I can learn from your example. And perhaps, some will realize that criticism is not as valuable a commodity, as they first imagined.

hushpuppy

Thanks for the kind words.

[Forester]

How would one diversify this model? Changing the lookback period paired with unemployment - it's not clear this would make much difference. The alternative would be a different economic indicator, or go 50-50 with a dumb longer moving average. The whipsaws won't feel so bad with a smaller slice of money.

Use unemployment as a filter
- lower incidence of whipsaws
- higher risk of following a steep market drop because unemployment is slow to shift

Use only a dumb moving average
- higher incidence of whipsaws
- assured of exiting as price deteriorates

[willthrill81]

How would one diversify this model? Changing the lookback period paired with unemployment - it's not clear this would make much difference. The alternative would be a different economic indicator, or go 50-50 with a dumb longer moving average. The whipsaws won't feel so bad with a smaller slice of money.

Use unemployment as a filter
- lower incidence of whipsaws
- higher risk of following a steep market drop because unemployment is slow to shift

Use only a dumb moving average
- higher incidence of whipsaws
- assured of exiting as price deteriorates

Actually, the 12 month moving average in the unemployment rate has been a leading indicator for every recession since WW2, with an average lead time of about 3.5 months.

Using multiple moving averages with different proportions of one's portfolio has been suggested before (e.g. 1/3 with 50 day moving average, 1/3 with 100 DMA, 1/3 with 200 DMA). Considering that the long-term results of most of these have been similar, I don't personally feel the need for their added complexity.

[michaeljc70]

Interesting. This isn't my approach; however, I appreciate your transparency and will grab my popcorn to follow your thread.

One question: How will you know when your plan is a success and by what metric? (You stated that one year's data is insufficient to gauge success.)

Livesoft asked the same question. I'll probably compare my results to both VTSAX and a combination of 85% VTSAX and 15% VBTLX since that's about the ratio that I expect to be in stocks (85% of the time) over the long-term.

Maybe I am misunderstanding your strategy, but can't you wind up being in a small cap fund or REIT for months (from your OP)? I don't think that VTSAX or VBTLX is an equivalent benchmark. Frankly, I don't know how you'd come up with one. Comparing to a benchmark that has very different risk levels is not really valuable.

[willthrill81]

Interesting. This isn't my approach; however, I appreciate your transparency and will grab my popcorn to follow your thread.

One question: How will you know when your plan is a success and by what metric? (You stated that one year's data is insufficient to gauge success.)

Livesoft asked the same question. I'll probably compare my results to both VTSAX and a combination of 85% VTSAX and 15% VBTLX since that's about the ratio that I expect to be in stocks (85% of the time) over the long-term.

Maybe I am misunderstanding your strategy, but can't you wind up being in a small cap fund or REIT for months (from your OP)? I don't think that VTSAX or VBTLX is an equivalent benchmark. Frankly, I don't know how you'd come up with one. Comparing to a benchmark that has very different risk levels is not really valuable.

I'm inclined to agree, but for the purposes of determining if my strategy was successful in achieving my objectives, I must compare it to something. If you have ideas for a better comparison, please feel free to share them.

[BlueEars]

The weakness of this trend following approach (and there is much to like) would be if unemployment did not show up soon enough to avoid a really really bad decline. If the unemployment gating function does not activate this would be a problem. In other words, just using a straight SP500 price approach adds in whipsaws but does provide a clear signal ahead of potentially many more months of declines.

I think that a really really bad sustained decline in the SP500 has not happened without an accompanying rise in unemployment. In the 1930's there was probably a good signal in unemployment i.e. unemployment did show up in the stats before too many declining months. Or if the stats were not available back then, theoretically one could have guessed at the signal. I have seen no monthly 1930's unemployment data.

In the October 1987 crash unemployment did not signal to get out and so the results were the same as buy/hold. Drawdowns were pretty large then but not of the 1930's sizes.

So what could happen that is really bad for stocks and not affect unemployment or at least not early enough to give a proper sell signal when combined with the equity pricing behavior?

[willthrill81]

The weakness of this trend following approach (and there is much to like) would be if unemployment did not show up soon enough to avoid a really really bad decline. If the unemployment gating function does not activate this would be a problem. In other words, just using a straight SP500 price approach adds in whipsaws but does provide a clear signal ahead of potentially many more months of declines.

I think that a really really bad sustained decline in the SP500 has not happened without an accompanying rise in unemployment. In the 1930's there was probably a good signal in unemployment i.e. unemployment did show up in the stats before too many declining months. Or if the stats were not available back then, theoretically one could have guessed at the signal. I have seen no monthly 1930's unemployment data.

In the October 1987 crash unemployment did not signal to get out and so the results were the same as buy/hold. Drawdowns were pretty large then but not of the 1930's sizes.

So what could happen that is really bad for stocks and not affect unemployment or at least not early enough to give a proper sell signal when combined with the equity pricing behavior?

I considered this issue quite a lot before deciding on this approach. It basically comes down to which con you'd rather accept: stocks potentially dropping substantially before or without an increase in the UER or dealing with much more frequent whipsaws. I chose the former.

[hdas]

I'm not at all concerned about a single whipsaw or even a series of them. I've studied the history of the markets and strategies like mine to know that that's inevitable, just as deep drawdowns are an inevitability of buy-and-hold (and might still happen to me to). That's a price that I'm willing to pay, just as many here are willing to forego the likely outperform of stocks for the stability of bonds.

Further, you could apply the same argument to those with value tilts and U.S.-only portfolios. Both are sure to underperform at some point in the future.

I'm assuming your system is still out of the market....so we can say the system is underperforming temporally by about 10% relative to the index, in 1 month since exiting. Since you have studied this extensively, can you share (putting numbers on the table):

1. How many times has this happen in the past according to your backtest?
2. What % of the total number of times the system got out of stocks does the above represent.
3. What are the likely outcomes going forward ?
4. What is the biggest 1 whipsaw the system ever experienced, and when was it?
5. What is the longest number of consecutive whipsaws the system experience and what was the cumulative under performance relative to the index?
6. When was the maximum spread of under-performance of the system vs index and by how much?

This would be very informative to all ppl following your thread. Thanks!!

[bgf]

I think whole thread is a collective fail by the Boglehead community. Why the deference and applause for a market timing scheme? Why aren't forum veterans pointing out that buying and selling of an entire portfolio based on some economic indicators is ill-considered?

Most folks who have put forth timing schemes like this been thumped with a collective BH response. Jack Bogle famously said he didn't know anybody could successfully time the Market and didn't know anybody who knew anybody who could successfully time the Market. That usually is the standard response when these threads are floated.

Willthrill's timing method may appear more data driven and less random, but it's still something we older experienced investors should discourage, if just for the sake of all the newbies reading along!

Probably because most of us like willthrill

The feeling's mutual.

But it probably won't hurt him too bad... He can afford to under-perform by 2%-4%, which is probably his worse case.

That's my assessment as well. I don't think that I'm likely to significantly underperform an 85/15 fixed AA over the long-term, which most here would consider to be aggressive anyway.

underperformance by 2-4% is massive. that's not like, 'hey, you know it wasn't SO bad.' because that really is truly absolutely unforgivably horrible over the long term.

[Forester]

We have to be careful not to compare such trend following models to a 100% equity benchmark.

https://alphaarchitect.com/2018/10/16/w ... following/

[willthrill81]

I'm not at all concerned about a single whipsaw or even a series of them. I've studied the history of the markets and strategies like mine to know that that's inevitable, just as deep drawdowns are an inevitability of buy-and-hold (and might still happen to me to). That's a price that I'm willing to pay, just as many here are willing to forego the likely outperform of stocks for the stability of bonds.

Further, you could apply the same argument to those with value tilts and U.S.-only portfolios. Both are sure to underperform at some point in the future.

I'm assuming your system is still out of the market....so we can say the system is underperforming temporally by about 10% relative to the index, in 1 month since exiting. Since you have studied this extensively, can you share (putting numbers on the table):

1. How many times has this happen in the past according to your backtest?
2. What % of the total number of times the system got out of stocks does the above represent.
3. What are the likely outcomes going forward ?
4. What is the biggest 1 whipsaw the system ever experienced, and when was it?
5. What is the longest number of consecutive whipsaws the system experience and what was the cumulative under performance relative to the index?
6. When was the maximum spread of under-performance of the system vs index and by how much?

This would be very informative to all ppl following your thread. Thanks!!

Yes, this event seems likely to be a whipsaw, but that cannot be known with certainty until we see what price I buy back in at.

The Philosophical Economics link I posted in the OP has much of the information you're asking about. I have not personally conducted an exhaustive backtest of the strategy.

[bloom2708]

Is a day like today a "Really Bad Day" when you are 0/100? Note: DOW up +1.5%, S&P and NASDAQ up 1.3 to 1.5%

It seems there are 10, 15 days like today no strategy can afford to be out of the market. I guess the big down days are "saved" by being out and turn into "Really Good Days".

That means you have 2 flavors of "Really Good Days" and "Really Bad Days". One set while "In" and one set while "Out".

[willthrill81]

Is a day like today a "Really Bad Day" when you are 0/100? Note: DOW up +1.5%, S&P and NASDAQ up 1.3 to 1.5%

It seems there are 10, 15 days like today no strategy can afford to be out of the market. I guess the big down days are "saved" by being out and turn into "Really Good Days".

That means you have 2 flavors of "Really Good Days" and "Really Bad Days". One set while "In" and one set while "Out".

Evaluating any strategy on the basis of such a short period of time is a practice rife with problems.

I knew full well when I decided to implement this strategy that whipsaws would happen. I'm much more concerned with avoiding a 50% loss than missing out on a 10% gain.

[bloom2708]

Is a day like today a "Really Bad Day" when you are 0/100? Note: DOW up +1.5%, S&P and NASDAQ up 1.3 to 1.5%

It seems there are 10, 15 days like today no strategy can afford to be out of the market. I guess the big down days are "saved" by being out and turn into "Really Good Days".

That means you have 2 flavors of "Really Good Days" and "Really Bad Days". One set while "In" and one set while "Out".

Evaluating any strategy on the basis of such a short period of time is a practice rife with problems.

I knew full well when I decided to implement this strategy that whipsaws would happen. I'm much more concerned with avoiding a 50% loss than missing out on a 10% gain.

I understand. Those small number of "Really Good Days" and a handful of individual stocks seem to account for a lot of the net gains during a year. Just curious if this was part of the thought process.

[willthrill81]

Is a day like today a "Really Bad Day" when you are 0/100? Note: DOW up +1.5%, S&P and NASDAQ up 1.3 to 1.5%

It seems there are 10, 15 days like today no strategy can afford to be out of the market. I guess the big down days are "saved" by being out and turn into "Really Good Days".

That means you have 2 flavors of "Really Good Days" and "Really Bad Days". One set while "In" and one set while "Out".

Evaluating any strategy on the basis of such a short period of time is a practice rife with problems.

I knew full well when I decided to implement this strategy that whipsaws would happen. I'm much more concerned with avoiding a 50% loss than missing out on a 10% gain.

I understand. Those small number of "Really Good Days" and a handful of individual stocks seem to account for a lot of the net gains during a year. Just curious if this was part of the thought process.

What's really interesting is that the "Really Good Days" tend to be clustered rather closely together with "Really Bad Days" during times when the market is in a downward trend. This is what is lost in the "If you missed the top X days in the market, your return would have been Y" idea that many promote.

[bloom2708]

Is a day like today a "Really Bad Day" when you are 0/100? Note: DOW up +1.5%, S&P and NASDAQ up 1.3 to 1.5%

It seems there are 10, 15 days like today no strategy can afford to be out of the market. I guess the big down days are "saved" by being out and turn into "Really Good Days".

That means you have 2 flavors of "Really Good Days" and "Really Bad Days". One set while "In" and one set while "Out".

Evaluating any strategy on the basis of such a short period of time is a practice rife with problems.

I knew full well when I decided to implement this strategy that whipsaws would happen. I'm much more concerned with avoiding a 50% loss than missing out on a 10% gain.

I understand. Those small number of "Really Good Days" and a handful of individual stocks seem to account for a lot of the net gains during a year. Just curious if this was part of the thought process.

What's really interesting is that the "Really Good Days" tend to be clustered rather closely together with "Really Bad Days" during times when the market is in a downward trend. This is what is lost in the "If you missed the top X days in the market, your return would have been Y" idea that many promote.

I appreciate your willingness to be your own test subject. Especially here. Missing January 2019 could ripple forward and be a multiple hundred thousand dollar ripple 20 years down the line. I have no idea what the next month or 6 months or 10 months will bring. Hopefully you are doing this with $100k and not $1 million. I guess you understand that. There is no undo button and thus is the reason this site exists. I should have no trepidation for you or your strategy, but I do. I will follow silently and hope you are on the right side of really good days more frequently than not.

[willthrill81]

Is a day like today a "Really Bad Day" when you are 0/100? Note: DOW up +1.5%, S&P and NASDAQ up 1.3 to 1.5%

It seems there are 10, 15 days like today no strategy can afford to be out of the market. I guess the big down days are "saved" by being out and turn into "Really Good Days".

That means you have 2 flavors of "Really Good Days" and "Really Bad Days". One set while "In" and one set while "Out".

Evaluating any strategy on the basis of such a short period of time is a practice rife with problems.

I knew full well when I decided to implement this strategy that whipsaws would happen. I'm much more concerned with avoiding a 50% loss than missing out on a 10% gain.

I understand. Those small number of "Really Good Days" and a handful of individual stocks seem to account for a lot of the net gains during a year. Just curious if this was part of the thought process.

What's really interesting is that the "Really Good Days" tend to be clustered rather closely together with "Really Bad Days" during times when the market is in a downward trend. This is what is lost in the "If you missed the top X days in the market, your return would have been Y" idea that many promote.

I appreciate your willingness to be your own test subject. Especially here. Missing January 2019 could ripple forward and be a multiple hundred thousand dollar ripple 20 years down the line. I have no idea what the next month or 6 months or 10 months will bring. Hopefully you are doing this with $100k and not $1 million. I guess you understand that. There is no undo button and thus is the reason this site exists. I should have no trepidation for you or your strategy, but I do. I will follow silently and hope you are on the right side of really good days more frequently than not.

Believe me, I understand that full well. The 100% equities person didn't get an "undo button" either for 2000-2009, whereas most trend following methods did fairly well during that period.

All roads carry risk, and I'm not at all convinced that my strategy is inherently riskier than buy-and-hold over the long-term.

[bloom2708]

Evaluating any strategy on the basis of such a short period of time is a practice rife with problems.

I knew full well when I decided to implement this strategy that whipsaws would happen. I'm much more concerned with avoiding a 50% loss than missing out on a 10% gain.

I understand. Those small number of "Really Good Days" and a handful of individual stocks seem to account for a lot of the net gains during a year. Just curious if this was part of the thought process.

What's really interesting is that the "Really Good Days" tend to be clustered rather closely together with "Really Bad Days" during times when the market is in a downward trend. This is what is lost in the "If you missed the top X days in the market, your return would have been Y" idea that many promote.

I appreciate your willingness to be your own test subject. Especially here. Missing January 2019 could ripple forward and be a multiple hundred thousand dollar ripple 20 years down the line. I have no idea what the next month or 6 months or 10 months will bring. Hopefully you are doing this with $100k and not $1 million. I guess you understand that. There is no undo button and thus is the reason this site exists. I should have no trepidation for you or your strategy, but I do. I will follow silently and hope you are on the right side of really good days more frequently than not.

Believe me, I understand that full well. The 100% equities person didn't get an "undo button" either for 2000-2009, whereas most trend following methods did fairly well during that period.

All roads carry risk, and I'm not at all convinced that my strategy is inherently riskier than buy-and-hold over the long-term.

As a "typical" Boglehead, I would never (at any age) recommend 100% equities. Well, maybe 18-22. So, I don't like the 100/0 argument. I worry you will put yourself in a box and have to stick with it to prove your point. It is probably what worries others too. If you thrash the market for 10 years and then write a book, sell millions of copies, then the strategy won't work at all for the next 10 years. Nobody knows nothing rules the day.

[willthrill81]

I understand. Those small number of "Really Good Days" and a handful of individual stocks seem to account for a lot of the net gains during a year. Just curious if this was part of the thought process.

What's really interesting is that the "Really Good Days" tend to be clustered rather closely together with "Really Bad Days" during times when the market is in a downward trend. This is what is lost in the "If you missed the top X days in the market, your return would have been Y" idea that many promote.

I appreciate your willingness to be your own test subject. Especially here. Missing January 2019 could ripple forward and be a multiple hundred thousand dollar ripple 20 years down the line. I have no idea what the next month or 6 months or 10 months will bring. Hopefully you are doing this with $100k and not $1 million. I guess you understand that. There is no undo button and thus is the reason this site exists. I should have no trepidation for you or your strategy, but I do. I will follow silently and hope you are on the right side of really good days more frequently than not.

Believe me, I understand that full well. The 100% equities person didn't get an "undo button" either for 2000-2009, whereas most trend following methods did fairly well during that period.

All roads carry risk, and I'm not at all convinced that my strategy is inherently riskier than buy-and-hold over the long-term.

As a "typical" Boglehead, I would never (at any age) recommend 100% equities. Well, maybe 18-22. So, I don't like the 100/0 argument. I worry you will put yourself in a box and have to stick with it to prove your point. It is probably what worries others too. If you thrash the market for 10 years and then write a book, sell millions of copies, then the strategy won't work at all for the next 10 years. Nobody knows nothing rules the day.

I appreciate your concern. I'm aware of the risks.

With your last statement, are you denying that the momentum effect exists?

[bloom2708]

What's really interesting is that the "Really Good Days" tend to be clustered rather closely together with "Really Bad Days" during times when the market is in a downward trend. This is what is lost in the "If you missed the top X days in the market, your return would have been Y" idea that many promote.

I appreciate your willingness to be your own test subject. Especially here. Missing January 2019 could ripple forward and be a multiple hundred thousand dollar ripple 20 years down the line. I have no idea what the next month or 6 months or 10 months will bring. Hopefully you are doing this with $100k and not $1 million. I guess you understand that. There is no undo button and thus is the reason this site exists. I should have no trepidation for you or your strategy, but I do. I will follow silently and hope you are on the right side of really good days more frequently than not.

Believe me, I understand that full well. The 100% equities person didn't get an "undo button" either for 2000-2009, whereas most trend following methods did fairly well during that period.

All roads carry risk, and I'm not at all convinced that my strategy is inherently riskier than buy-and-hold over the long-term.

As a "typical" Boglehead, I would never (at any age) recommend 100% equities. Well, maybe 18-22. So, I don't like the 100/0 argument. I worry you will put yourself in a box and have to stick with it to prove your point. It is probably what worries others too. If you thrash the market for 10 years and then write a book, sell millions of copies, then the strategy won't work at all for the next 10 years. Nobody knows nothing rules the day.

I appreciate your concern. I'm aware of the risks.

With your last statement, are you denying that the momentum effect exists?

I am not denying the momentum effect exists. Detecting it, timing it, understanding hidden and false signals, reacting too fast or to slow, rapid changes in momentum, how far along before it becomes visible, those are much harder than looking backwards and seeing it in plain sight.

[willthrill81]

I appreciate your willingness to be your own test subject. Especially here. Missing January 2019 could ripple forward and be a multiple hundred thousand dollar ripple 20 years down the line. I have no idea what the next month or 6 months or 10 months will bring. Hopefully you are doing this with $100k and not $1 million. I guess you understand that. There is no undo button and thus is the reason this site exists. I should have no trepidation for you or your strategy, but I do. I will follow silently and hope you are on the right side of really good days more frequently than not.

Believe me, I understand that full well. The 100% equities person didn't get an "undo button" either for 2000-2009, whereas most trend following methods did fairly well during that period.

All roads carry risk, and I'm not at all convinced that my strategy is inherently riskier than buy-and-hold over the long-term.

As a "typical" Boglehead, I would never (at any age) recommend 100% equities. Well, maybe 18-22. So, I don't like the 100/0 argument. I worry you will put yourself in a box and have to stick with it to prove your point. It is probably what worries others too. If you thrash the market for 10 years and then write a book, sell millions of copies, then the strategy won't work at all for the next 10 years. Nobody knows nothing rules the day.

I appreciate your concern. I'm aware of the risks.

With your last statement, are you denying that the momentum effect exists?

I am not denying the momentum effect exists. Detecting it, timing it, understanding hidden and false signals, reacting too fast or to slow, rapid changes in momentum, how far along before it becomes visible, those are much harder than looking backwards and seeing it in plain sight.

Absolutely. That's why I know that will certainly underperform the market at times, perhaps for stretches lasting years.

[HomerJ]

50/50 and you never have to second-guess yourself

50/50 also means I leave a lot of potential gains on the table because I'm too conservative, but it also strongly mutes any losses I might experience during the downturns.

If the stock market is going up, I own a good chunk, and I'm happy.

If the stock market is going down, I have a good chunk in safe bonds, and I'm happy.

I'm never wrong, and I never have to wonder if I made a mistake.

[H-Town]

50/50 and you never have to second-guess yourself

True.

Anywhere from 40/60 to 60/40 should allow portfolio to avoid huge draw-down.

[diy60]

When my system calls for me to move back into stocks, I'll post that here.

WT, thanks for this thread. I've been a diy investor for 35+ years and have tried multiple strategies, none of which worked for me. Or at least, any that I stuck with. I'm not a robot, and I found my self constantly justifying inaction or action, never quite sticking to absolute rules. Last 20 years has been pretty much buy the standard stuff and don't look at it.

I don't have the numbers in front of me, but I'm thinking the ^GSPC crossed above both the 200 day and 7 MMA. Q: When during the month do you pull the trigger? Sorry if you already have stated that info somewhere in this multi page thread. Best of luck, and please keep the thread updated.

[Always passive]

Do I then understand that you do not worry about allocation of stocks and bonds, you are either 100% in stocks (picking that with the highest momentum) or 100% in fixed income when your "get out of the market" signal kicks in?

[willthrill81]

When my system calls for me to move back into stocks, I'll post that here.

WT, thanks for this thread. I've been a diy investor for 35+ years and have tried multiple strategies, none of which worked for me. Or at least, any that I stuck with. I'm not a robot, and I found my self constantly justifying inaction or action, never quite sticking to absolute rules. Last 20 years has been pretty much buy the standard stuff and don't look at it.

I don't have the numbers in front of me, but I'm thinking the ^GSPC crossed above both the 200 day and 7 MMA. Q: When during the month do you pull the trigger? Sorry if you already have stated that info somewhere in this multi page thread. Best of luck, and please keep the thread updated.

I trade on the first Friday of the month, when UER data is released.

[willthrill81]

Do I then understand that you do not worry about allocation of stocks and bonds, you are either 100% in stocks (picking that with the highest momentum) or 100% in fixed income when your "get out of the market" signal kicks in?

That's correct.

[Always passive]

Why do you use a moving average to stay or not in the market, and momentum to pick the asset and not Gary Antonacci's dual momentum?