Simple Timing system

[RockOn]

T A A N S A A F L.

Thousands of stock pickers and "technical analysts" (including Zweig) make their living convincing people that it is possible.

I could build you a back-tested model that does much better than Zweig's.

IMO, the most trustworthy and realistic way of reducing portfolio risk (measured as volatility) while increasing returns is through use of modern portfolio theory. The Basics of Investing and Portfolio Theory A lot of responsible folks have looked into it, and it has academic backing. It also works in the real world.

That debate will continue forever I think. There are also many who say modern portfolio theory is completely bunk. It's easy to dismiss either side based on your point of few. "Works in the real world" is impossible to measure because the future is always unlike the past. So on it goes.....

You might be right that risk cannot be reduced while getting similar returns. That might be true overall but it is not true during certain periods with certain methods, this study proves that. I really am not sure because I see both sides make valid arguements. I'll lean toward your view, but in times like this (extreme uncertainty I would call it), I waiver some. There might be times when the risks are not worth faithfully following a theory.

 

[al_bundy]

An observation that will probably come as no surprise to those here: Trading options (selling or buying either puts or calls) is a zero sum game. By definition, for every person that makes a dollar, someone loses a dollar. For every person on one end of a trade wo thinks he knows what will happen, there's someone else on the other end who is sure he knows. Add in substantial transaction costs and it's clear who the long-term winner always is.

Owning actual stocks (or MFs/ETFs, etc) is different--it is possible for everyone to make money.

that's only if you think of trading options like stocks, they are used extensively to hedge risk and if you read a basic options trading book or what any broker writes in their how to trade options page there are a few strategies where you can go long and short at the same time and make money on which makes more profit

i've sold call options before just to play with them, didn't lose anything. someone paid me to buy my stock at a specified price. and he or she got it

 

[LOL!]

I'm not sure if this thread is still on topic or not. A simple timing system is run by James Stewart (I searched this thread for his name using the simple search tool on the forum, so not sure if it was mentioned in this thread already). His method is reminiscent of the OP.
http://www.smartmoney.com/common-sense/
I probably got this wrong, but his method is sell on a 25% rise in stock and buy after a 10% pull-back. I forget which index he uses, but I think it's the NASDAQ.

 

[Zathras]

Well, at least then you are buying low, selling high.
The original idea in this thread (which I think has been dismissed by the OP (please correct me if I am wrong), was to sell low and buy high

 

[RockOn]

Well, at least then you are buying low, selling high.
The original idea in this thread (which I think has been dismissed by the OP (please correct me if I am wrong), was to sell low and buy high

The thread has drifted a bit, but to summarize. I have not dismissed the idea, it still might be a very good idea for this market, but the significant weaknesses of the approach have been identified.

It's not the holy grail of investment timing. If the bull market returns it will be easy to say why bother. If the SP500 falls another 30%, you may wish you had tried something like this.

To be clear on how simple my idea is...let's assume the level was set at Dow 11,400. I'd be 100% long right now as of Thursday. If the DOW falls below 11,400 I'd sell out 100% and wait for it to go above 11,400 before getting 100% long again. I could not miss out on another bull run that way, yet if the bear continues I'll be completely out. The problem is every time it goes above 11,400 and then below 11,400 before the next new bull market resumes, I'd have a whipsaw loss. How many times will that happen and what will those losses amount to? That's the risk and nobody knows ahead of time. At times like this it seems like it could be less risk than hoping for a bull market soon and possibly end up being very wrong about that.

 

[guest]

Guest, would you let us know when you decide to get back in the market? You seem to be pretty good at it.

or really lucky would probably be more accurate! There have been times when I feel as others have shared...."By my actions, I alone have changed the direction of the market!" hahaha

I like where I am right now....not losing. Soon I will be complaining that I'm not making enough followed by......

 

[Hamlet]

So your system would have you sell today. You've been whip-sawed once now.

It will be interesting to watch this.

To be clear on how simple my idea is...let's assume the level was set at Dow 11,400. I'd be 100% long right now as of Thursday. If the DOW falls below 11,400 I'd sell out 100% and wait for it to go above 11,400 before getting 100% long again. I could not miss out on another bull run that way, yet if the bear continues I'll be completely out. The problem is every time it goes above 11,400 and then below 11,400 before the next new bull market resumes, I'd have a whipsaw loss. How many times will that happen and what will those losses amount to? That's the risk and nobody knows ahead of time. At times like this it seems like it could be less risk than hoping for a bull market soon and possibly end up being very wrong about that.

 

[RockOn]

So your system would have you sell today. You've been whip-sawed once now.

It will be interesting to watch this.

I was going to post that whipsaw tonight but you beat me to it. With a 11,400 price level you are correct, I'd be down about 0.9% so far. But I would be sleeping very well tonight, knowing that if this was a suckers rally, I am safely 100% out.

Yes, it will interesting to watch what happens.

 

[ERD50]

Yes, it will interesting to watch what happens.

I agree - I hope we do follow it and post here. An hopefully, people don't think that that means we want to 'beat up' on RockOn if it fails, I simply want to learn from the idea.

While I am very skeptical that anything like this can 'work', I don't want to dismiss it out-of-hand. I think there is some merit to what RockOn proposed, but I also suspect the whipsaw/gap issues will doom it, but how do we know unless we track it?

We might learn something, even if that something is, 'this didn't work'.

-ERD50

 

[RockOn]

I agree - I hope we do follow it and post here. An hopefully, people don't think that that means we want to 'beat up' on RockOn if it fails, I simply want to learn from the idea.

While I am very skeptical that anything like this can 'work', I don't want to dismiss it out-of-hand. I think there is some merit to what RockOn proposed, but I also suspect the whipsaw/gap issues will doom it, but how do we know unless we track it?

We might learn something, even if that something is, 'this didn't work'.

-ERD50

We will see, it looks there could be another buy tonight. 11,400 was just picked out of the air, but it is as good as anything for watching this. If I were doing it for real, I'd look at support/resistance lines to pick my number. Of course, that's if one believes they have any value.

I don't care if I get beat up for this if it fails, it might but it might not! Beating up on me is fun anyway. I wouldn't let a few whipsaws bother me yet. After about 5 or 10, I'd be starting to get concerned.

If the market would drop significantly I would lower my number which would be good. We'll see if I get that chance. If the market rises enough, I may raise my number so the next whipsaw would be free. I'll post my revisions and the signals for you until it becomes boring. If the bear market is over, there will not be much to watch.

 

[FIRE'd@51]

I agree - I hope we do follow it and post here. An hopefully, people don't think that that means we want to 'beat up' on RockOn if it fails, I simply want to learn from the idea.

While I am very skeptical that anything like this can 'work', I don't want to dismiss it out-of-hand. I think there is some merit to what RockOn proposed, but I also suspect the whipsaw/gap issues will doom it, but how do we know unless we track it?

We might learn something, even if that something is, 'this didn't work'.

-ERD50

What's to follow? RockOn has proposed a simple trading rule. He trades once a day (if necessary) based upon the Dow's close. If it's below his strike price (e.g. 11,400), he is 100% cash - if it's above, he is 100% stock. We have years of historical data for the Dow. Just download the historical Dow data into a spreadsheet, and apply the trading rule. It couldn't be more simple. The cost relative to his desired payoff (i.e., the number of whipsaw losses) is how he pays for (i.e., the premium of) his "synthetic option".

This will be the theoretical best he can do, since there will be additional logistical problems associated with trading at the closing price, which will increase his transaction costs. Most mutual fund companies require that orders be placed at least 10 minutes before the close. Also, most have blackout periods after the sale (they don't want people dynamically trading their fund which runs up the expenses for all shareholders). This means that other transaction costs will be higher than RockOn claims. To actually implement the strategy will require trading ETF's (or futures) which involve commissions and bid/ask spreads, and, in the case of futures, variation margin. This will increase the whipsaw cost.

 

[RockOn]

What's to follow? RockOn has proposed a simple trading rule. He trades once a day (if necessary) based upon the Dow's close. If it's below his strike price (e.g. 11,400), he is 100% cash - if it's above, he is 100% stock. We have years of historical data for the Dow. Just download the historical Dow data into a spreadsheet, and apply the trading rule. It couldn't be more simple. The cost relative to his desired payoff (i.e., the number of whipsaw losses) is how he pays for (i.e., the premium of) his "synthetic option".

This will be the theoretical best he can do, since there will be additional logistical problems associated with trading at the closing price, which will increase his transaction costs. Most mutual fund companies require that orders be placed at least 10 minutes before the close. Also, most have blackout periods after the sale (they don't want people dynamically trading their fund which runs up the expenses for all shareholders). This means that other transaction costs will be higher than RockOn claims. To actually implement the strategy will require trading ETF's (or futures) which involve commissions and bid/ask spreads, and, in the case of futures, variation margin. This will increase the whipsaw cost.

I mostly agree. Yes there are some trading costs, I did not say there were none, even though I suppose it could have been implied. I could buy and sell Profunds or Rydex index funds at any broker for very low cost without trading limits. I could even go 2X leverage if I wanted too. Those funds do have large fees, so granted there are some costs to use them. As far as the cutoff time for trading, you are correct. That doesn't come into play that often, but it certainly could.

As far as using historical data, I don't think that applies because I am proposing to use during times of crisis management, not for regularly trading markets. (If one could avoid the whipsaws and gaps, it could be used all the time.) I know it is not clear when crisis times are, but when banks are going down and foreclosures are soaring, I'd say this is one of those times.

 

[free4now]

In my early investing career, I traded a system very much like the OP's proposal... I sold when the price dropped below the 40 week moving average, and bought when the price went above the 40 WAR.

It was the whipsaws that finally got me to stop. Basically you are making a tradeoff: You are exchanging the rare possibility of whipsaws taking you down bad for a most likely possibility of slightly lower volatility and same or slightly lower returns.

It is kind of like getting in your car and driving on the highway: Most of the time it's going to work out great... but every now and then it goes horribly wrong and people die.

Another thing that got me off of trading and into buy and hold was the bid/ask spread. I don't think this has been mentioned here, but if you are trading equities there will always be a small difference the price you can get selling and what you can buy for. Every time you make an in and out trade cycle, you effectively lost that amount of money.

The argument that the whipsaws are not so much of a concern because you're not going to use this strategy very often? That's like saying you're not worried about how your tires perform in the rain because it doesn't rain very much. Or that because the downside risks are so great in the market right now that it's worthwhile for you to add more downside risks yourself?

Another thing to consider is that with a trading system like the OP proposes, any mistake will likely work against you; because you are selling when prices are falling, and buying when prices are rising, if you delay any of your trades for any reason the delay will very likely work out strongly against you.

And there are times when whipsaws happen; prices can cross above and below a particular price point hundreds of times in a short period of time. The only thing that you can do is delay your trades on one end or the other, which will cost you.

 

[haha]

Systems like this can work if you have an opinion on market direction. Say you wish to go long because you feel that stocks are strongly undervalued or whatever other fundamental reason you may have for expecting a large upward move, and all you want is timing and loss protection.

You trade your filter, with parameters that are suitable for the given index or stock or futures position. The rule might be an upside penetration by a certain %, or a penetration that is maintained for a week, or whatever you like. Use another rule for getting out. It only works if you are right on direction and size of move, but it is a way to probe the upside, or downside using inverse funds, with controlled risk.

Ha

 

[ERD50]

ETFs like SPY trade at very low spread, low commissions and no trade limit. Market's closed now, but I'll look again Monday, pretty slim IIRC.

-ERD50

 

[Hamlet]

The thing is, these "crisis times" come up with a fair amount of frequency. I count three times in the last ten years. The "Asian contagion, LTCM" downswing of 1998, the "tech meltdown-9/11" of 2000-2003, and now the "housing bubble". In the decade before that, there was the S&L crisis, Black Monday, and the banking crisis of the early 90s.

We've seen a fair amount of the earnings for this quarter. The banks that stuck to decent standards (USB, WFC, JPM) are doing ok. Others are not doing well, and may have to raise more capital. A very few have gone under.

Honestly, if you never watched any financial channel like CNBC, would you even know there was a problem?

We've had some minor job loses from the housing downturn, but housing has always been a boom/bust business, we just forgot that for 10 years.

Gas is expensive, but most people only buy about a tank a week.

I just don't think this qualifies as a crisis yet. The 70s were a crisis. So far this is just noise.

As far as using historical data, I don't think that applies because I am proposing to use during times of crisis management, not for regularly trading markets. (If one could avoid the whipsaws and gaps, it could be used all the time.) I know it is not clear when crisis times are, but when banks are going down and foreclosures are soaring, I'd say this is one of those times.

 

[RockOn]

The thing is, these "crisis times" come up with a fair amount of frequency. I count three times in the last ten years. The "Asian contagion, LTCM" downswing of 1998, the "tech meltdown-9/11" of 2000-2003, and now the "housing bubble". In the decade before that, there was the S&L crisis, Black Monday, and the banking crisis of the early 90s.

We've seen a fair amount of the earnings for this quarter. The banks that stuck to decent standards (USB, WFC, JPM) are doing ok. Others are not doing well, and may have to raise more capital. A very few have gone under.

Honestly, if you never watched any financial channel like CNBC, would you even know there was a problem?

We've had some minor job loses from the housing downturn, but housing has always been a boom/bust business, we just forgot that for 10 years.

Gas is expensive, but most people only buy about a tank a week.

I just don't think this qualifies as a crisis yet. The 70s were a crisis. So far this is just noise.

Not a crisis yet....ok.........I'd agree there has not been a meltdown yet........but not a crisis yet.....ok.......We each see things differently I suppose

 

[RockOn]

ETFs like SPY trade at very low spread, low commissions and no trade limit. Market's closed now, but I'll look again Monday, pretty slim IIRC.

-ERD50

Those would work pretty well, the whipsaw losses could be small if the market is watched closely intraday. One could get several whipsaws in one day though. Overnight gap openings could also be a problem.

Instead of going long on these signals another option would be to buy 2X leverage short funds on a close below the number to fully hedge. Or there are many ways to use options.

The bottom line is how much risk do you think there is. If you think there could be a significant selloff in the near future some type of timing might make sense at times like this. The idea presented was the simplest of methods, more complicated methods may not actually be any better. For market timing, the simplest ideas tend to work best. Timing systems don't always work, they do work at times though.

We'll watch this for awhile. In several months we'll be able to see if the whipsaws were a price worth paying for limiting downside while still getting in on the upside.

As some have said, this isn't for everyone. If you cannot trade in and out because of fund restrictions or large commisions, it's really not a plan for you. I could do it with little cost, maybe you could, maybe not. There are some costs.

 

[ERD50]

Those would work pretty well, the whipsaw losses could be small if the market is watched closely intraday. One could get several whipsaws in one day though. Overnight gap openings could also be a problem.

I may have lost a detail along the way. You picked the DOW, 11,400 number; sell if it drops below, buy to hold if it goes above, right? But are those Market close #'s? I thought so, then you talked about intra-day (maybe this was mutual funds versus ETFs?).

If you are saying ETFs could be bought/sold intra-day, then you need to pick some hysteresis to make that practical. Some sort of 'banding' around 11,400, or you could end up trading 10x in a day if it was flittering right at 11,400 for a few hours. I count somewhere between 10 and 20 trades on this one-day chart, not enough resolution to be certain.

If you band that at 11,410 buy, 11390 sell, you would have six trades that day, and a 20 pnt loss on each buy/sell trip, so 60 points lost. On a day the DOW was up 220 from the prev day, and up from the open (according to their liste open price, can't see it on the chart).

Quotes for ^DJX - Yahoo! Finance

Maybe that is made up other days? I don't know.

-ERD50

 

[samclem]

The bad news is that you'd slowly go broke from transaction costs.

The good news is that you could count on getting a very nice Christmas card from your broker. (If E.Trade, they'd send you a very heartwarming autogenerated email)

 

[RockOn]

ERD50

You are taking this a little farther than I intended it, it is no longer simple. When I started the idea I was talking on a closing basis, not intraday.

But if you want to take it that far, even if I traded 20 times in a day and paid a commission each time, and lost 60 points in a day, I still might be better off than losing 30% of my assets if the market were to eventually tank. If I have a million dollars, that would be a $300k loss. That would pay a lot of $30 commissions. If I could trade intraday at a very low bid/ask and was on the button each time it crosses my number, the losses could be minimized somewhat. (If I were a sophisticated hedge fund I could probably figure out a way to keep the whipsaw costs down with futures/options and the like.) The odds are that a market trading on crisis type news, will not likely hang around my number for long. That is why I suggest it only for markets like this. If it did hang around my number, that would not be good but I still eventually would be on the right side of the market. There are no guarantees.

 

[ERD50]

OK, keep it simple and just trade based on market close.

But, you are saying to do this only when you view the market is particularly unstable, so that you might avoid a 30% (or whatever ) loss. Well, that is assuming that you are reasonably correct in your ability to predict times when the market is readty to diverge significantly. This might be such a time, or it might just move along in a 10% range for a long time.

So, if you choose to do this based on your judgment, instead of all the time, you risk absorbing the comm, fees, gaps, an whipsaws for no benefit sometimes, and you may still miss a 30% decline that you did not anticipate.

So unless you can include an objective definition of when to use this approach and when not to, you can't really measure it's success factor. Unless you just stay in it all the time. Otherwise all you'll have is (maybe), 'it works when it works, and it doesn't when it doesn't'.


edit/add: I'm not saying that to be negative, it's just the reality of the thing.

-ERD50

 

[free4now]

This all reminds me of a sentiment that I hear from my mother sometimes: "I'd like to invest prudently, but I'm so far behind (or the market is so bad) that will never get me where I need to be so I need to make investments that are riskier than I'm comfortable with".

There's a sense that we can wave our hands at obvious risks if it seems urgent or necessary. What brought this up was RockOn saying that losing 0.6% per normal day on transaction/whipsaw costs might be worthwhile. Sorry, but I just can't buy that. Any strategy that squeezes you that fast will bankrupt you in half a year. Yes, you could lose 30% by not following the strategy, but by following the strategy you will lose more than 30% in 60 days just on frictional costs.

 

[samclem]

Yes, you could lose 30% by not following the strategy, but by following the strategy you will lose more than 30% in 60 days just on frictional costs.

Well, I think we're all missing the point--you only use this system at certain times--when the market is "like this." When the market isn't "like this" you use a different set of rules (TBD). Or you buy an annuity and get a very juicy return. We already know how to maximize the return on any annuity--just assume that you'll live 6 years longer than the insurance tables say those of our age normally live--simple!

ERD50 is a very patient guy.

 

[Hamlet]

Man, the market missed triggering a buy for RockOn's system by less than 3 points!