19 ms·

Buy and hold forever is generally great advice, until it isn't. I'd much rather not have 30%+ of my life savings wiped out in a single recession with the rest

Buy and hold forever is generally great advice, until it isn't. I'd much rather not have 30%+ of my life savings wiped out in a single recession with the rest of the market. In my humble opinion, thoroughly backtested trading algorithms based on hard statistics is the only rational way to participate in the market. The market is fundamentally an irrational entity, and investors should be looking for ways to systematically protect themselves from the market's irrational mood swings. Check out Quantopian: https://www.quantopian.com/home https://www.quantopian.com/home Even a trivial moving average crossover algorithm like the one I currently have deployed can give you surprisingly robust protection against severe downswings, and outperform the overall market significantly over the long term despite the small losses caused by false positives. http://i.imgur.com/ZhN0QIp.png http://i.imgur.com/ZhN0QIp.png

frinxor 11y ago
you should start a hedge fund, and charge 2.5%! joking aside, i've played with quantopian way back, and think its a fun/cool concept
solve 11y ago
Can you show the forward performance, starting from the date when you chose those 4 parameters - 165.12, 172.44, 164.23, 156.02? As you know, backtested results with carefully tuned magic numbers can potentially mean nothing.
7Figures2Commas 11y ago
Your overall point is not a bad one, but investors don't need "thoroughly backtested trading algorithms" to protect themselves. Frankly, many of the people who use "thoroughly backtested trading algorithms" don't do as well as they supposedly should. It is, on the other hand, entirely possible for an average investor to learn basic technical analysis concepts and apply them visually to charts. At a minimum, for instance, if you can draw a trend line on a price chart and identify when price breaks important trend lines, you can easily avoid losses from major declines without having to write a single line of code or spend more than 5-10 minutes a day checking on your portfolio. A decent book in this vein is The Visual Investor[1]. [1] http://www.amazon.com/The-Visual-Investor-Market-Trends/dp/0470382058/ http://www.amazon.com/The-Visual-Investor-Market-Trends/dp/0...
- pinky1417 11y ago
  There's not a shred of evidence that technical analysis is anything but a charlatan's tool - same as trying to time the market. Technical analysis is bound to yield only losses for the investor and fees for the broker. The evidence points strongly against technical analysis as a method for building wealth - . And buying during upswings / selling during declines guarantees you lose. Long-term value investing performs best, low cost index funds perform well, Wealthfront will underperform due to fees, and technical analysis will turn even a billionaire into a millionaire.
  - 7Figures2Commas 11y ago
    There is plenty of charlatanism in the world of technical analysis, just as there is in the world of fundamental analysis. But you're missing a key point: "technical analysis" is not monolithic. There isn't a single set of rules and indicators, all applied the same way over the same time periods. Most investors (and traders) lose money because of lack of discipline, poor money management, etc. This is true whether technical analysis is used or not. Instead of creating a straw man argument, let's discuss something simple: trend lines. And let's discuss the most fundamental truth about trend lines: it is absolutely impossible for the current trend to reverse without key trend lines being broken. Don't take my word for it. Pull up a chart of the S&P 500. The bull market that ended in late 2007 was easily identified by the breaking of long-term trend lines. And the bear market that ended in 2009 was easily identified by the breaking of trend lines as well. Just by use of trend lines alone (and no other technical indicators) you could have identified key turning points in the market. How you act upon this information, if at all, is your choice. But that has nothing to do with technical analysis. That's investment/trading strategy.
    - forgetsusername 11y ago
      >The bull market that ended in late 2007 was easily identified by the breaking of long-term trend lines. And the bear market that ended in 2009 was easily identified by the breaking of trend lines as well. Trends are broken in periods surrounding major market corrections, indeed. How many false positives are there in between, in which you're turning over your portfolio for no reason?
gaius 11y ago
thoroughly backtested trading algorithms based on hard statistics is the only rational way to participate in the market. The market is fundamentally an irrational entity You do realize those two statements contradict each other?
- iaw 11y ago
  What's also funny to me is that 'thoroughly backtested trading algorithms based on hard statistics' will work consistently until they fail miserably. During a recession event the 30% loss isn't real until you close your position, with trading algorithms the time to hold is so low that any losses are very quickly realized.
gdubs 11y ago
If you're invested in an index fund, wouldn't the only way to realize the loss during a recession be to sell? Anyone who stayed the course after 2008 would have seen record gains in the years following.
- sadface 11y ago
  You're right, but it would still be better to realize a small loss early in the crash and and buy back in at the bottom than just ride the whole thing out. Of course, that kind of timing is pretty much impossible so everyone suggests you don't do it.
- minnend 11y ago
  Yes, but consider the "accumulation phase" (pre-retirement) vs. the "distribution phase" (retirement). At some point, you need to sell part of your portfolio to to pay your bills []. Check out "sequence of returns risks". Or buy annuities, setup a bond ladder, or hold dividend stocks, but these have their own, related issues. Early in the accumulation phase, however, I completely agree with you -- you're making the central argument for a long-term buy-and-hold strategy.
tosseraccount 11y ago
"backtested trading algorithms" Your algorithm can predict the past. Congratulations.
- calgaryeng 11y ago
  I don't think you know how these work
  - ScottBurson 11y ago
    tosseraccount has an excellent point. Overfitting is a huge danger. In the end only trading the live market for some time will tell you whether you have a profitable strategy.
brobinson 11y ago
Your algorithm will likely not stand up to Walk Forward Analysis/Optimization/Testing: https://en.wikipedia.org/wiki/Walk_forward_optimization https://en.wikipedia.org/wiki/Walk_forward_optimization I recommend reading "The Evaluation and Optimization of Trading Strategies" by Robert Pardo.
- solve 11y ago
  It's really the most fundamental rule for every type of machine learning task -- there should be both a (1) test set and (2) a separate holdout validation set. No matter if the task is classification, sequence prediction, regression, manifold learning, clustering, whatever. Really surprising how so few amateur traders understand this fundamental principal. Scientists too.
MaxScheiber 11y ago
> Buy and hold forever is generally great advice, until it isn't. That's a truism. I could say the same thing about algorithmic trading. The fact of the matter is that the median hedge fund manager has negative alpha. It is extremely difficult to pick winners via any strategy. In fact, it's so difficult that for people who can do it, the resulting lifestyle and compensation is so great that the asset management shops that they work out have almost zero attrition. Moreover, the whole point of passive investment is that the odds of that kind of a market-wide crash affecting your money over a 30-year timespan are infinitesimal. If the market actually were to devolve like that, nobody would make money in the market, regardless of strategy. > Investors should be looking for ways to systematically protect themselves from the market's irrational mood swings. Yes, it's called diversification. I won't harp on the backtesting point, but what I will say is this: I'm sure that you understand the risk-reward tradeoff. It would be impossible to participate in any sort of finance and not know this concept. Quite frankly, the probability of 30%+ of your life savings being wiped out in an index fund is unbelievably low. You eat some return at the expense of lower volatility, but we're talking about a 401(k) here. We want that kind of expected stability.
- tosseraccount 11y ago
  "the probability of 30%+ of your life savings being wiped out in an index fund is unbelievably low" The crash of 69 : 36% The crash of 73: 48% The crash of 87: 33% The dotcom crash 2000: 50% Housing Bubble 2007: 56% [ source : http://traderhq.com/illustrated-history-every-s-p-500-bear-market/ http://traderhq.com/illustrated-history-every-s-p-500-bear-m... ] The odds that at some point a portfolio of S&P 500 stocks will be 30% off a recent high is actually very high. Wars. Panics. Depressions. Plagues. Recessions. Revolutions. They still happen.
  - jordan0day 11y ago
    I don't know much about the market, and know even less about actual investing strategy-ish stuff, but what does the loss look like five years later? I guess what I mean is, for an "index fund", crashes don't seem to be... permanent? Perhaps that's small comfort for someone who's set to retire the month after one of these 30% crashes, but I guess that's what the whole "invest more conservatively as you get older" thing is supposed to help with?
nicholas73 11y ago
The only quibble I have is that taxes would outweigh the advantages here. You either do this in a tax free account, or your trading needs to be a lot better!

frinxor 11y ago

you should start a hedge fund, and charge 2.5%! joking aside, i've played with quantopian way back, and think its a fun/cool concept

solve 11y ago

Can you show the forward performance, starting from the date when you chose those 4 parameters - 165.12, 172.44, 164.23, 156.02? As you know, backtested results with carefully tuned magic numbers can potentially mean nothing.

7Figures2Commas 11y ago

Your overall point is not a bad one, but investors don't need "thoroughly backtested trading algorithms" to protect themselves. Frankly, many of the people who use "thoroughly backtested trading algorithms" don't do as well as they supposedly should. It is, on the other hand, entirely possible for an average investor to learn basic technical analysis concepts and apply them visually to charts. At a minimum, for instance, if you can draw a trend line on a price chart and identify when price breaks important trend lines, you can easily avoid losses from major declines without having to write a single line of code or spend more than 5-10 minutes a day checking on your portfolio. A decent book in this vein is The Visual Investor[1]. [1] http://www.amazon.com/The-Visual-Investor-Market-Trends/dp/0470382058/ http://www.amazon.com/The-Visual-Investor-Market-Trends/dp/0...

pinky1417 11y ago
There's not a shred of evidence that technical analysis is anything but a charlatan's tool - same as trying to time the market. Technical analysis is bound to yield only losses for the investor and fees for the broker. The evidence points strongly against technical analysis as a method for building wealth - . And buying during upswings / selling during declines guarantees you lose. Long-term value investing performs best, low cost index funds perform well, Wealthfront will underperform due to fees, and technical analysis will turn even a billionaire into a millionaire.
- 7Figures2Commas 11y ago
  There is plenty of charlatanism in the world of technical analysis, just as there is in the world of fundamental analysis. But you're missing a key point: "technical analysis" is not monolithic. There isn't a single set of rules and indicators, all applied the same way over the same time periods. Most investors (and traders) lose money because of lack of discipline, poor money management, etc. This is true whether technical analysis is used or not. Instead of creating a straw man argument, let's discuss something simple: trend lines. And let's discuss the most fundamental truth about trend lines: it is absolutely impossible for the current trend to reverse without key trend lines being broken. Don't take my word for it. Pull up a chart of the S&P 500. The bull market that ended in late 2007 was easily identified by the breaking of long-term trend lines. And the bear market that ended in 2009 was easily identified by the breaking of trend lines as well. Just by use of trend lines alone (and no other technical indicators) you could have identified key turning points in the market. How you act upon this information, if at all, is your choice. But that has nothing to do with technical analysis. That's investment/trading strategy.
  - forgetsusername 11y ago
    >The bull market that ended in late 2007 was easily identified by the breaking of long-term trend lines. And the bear market that ended in 2009 was easily identified by the breaking of trend lines as well. Trends are broken in periods surrounding major market corrections, indeed. How many false positives are there in between, in which you're turning over your portfolio for no reason?

gaius 11y ago

thoroughly backtested trading algorithms based on hard statistics is the only rational way to participate in the market. The market is fundamentally an irrational entity You do realize those two statements contradict each other?

iaw 11y ago
What's also funny to me is that 'thoroughly backtested trading algorithms based on hard statistics' will work consistently until they fail miserably. During a recession event the 30% loss isn't real until you close your position, with trading algorithms the time to hold is so low that any losses are very quickly realized.

gdubs 11y ago

If you're invested in an index fund, wouldn't the only way to realize the loss during a recession be to sell? Anyone who stayed the course after 2008 would have seen record gains in the years following.

sadface 11y ago
You're right, but it would still be better to realize a small loss early in the crash and and buy back in at the bottom than just ride the whole thing out. Of course, that kind of timing is pretty much impossible so everyone suggests you don't do it.
minnend 11y ago
Yes, but consider the "accumulation phase" (pre-retirement) vs. the "distribution phase" (retirement). At some point, you need to sell part of your portfolio to to pay your bills []. Check out "sequence of returns risks". Or buy annuities, setup a bond ladder, or hold dividend stocks, but these have their own, related issues. Early in the accumulation phase, however, I completely agree with you -- you're making the central argument for a long-term buy-and-hold strategy.

tosseraccount 11y ago

"backtested trading algorithms" Your algorithm can predict the past. Congratulations.

calgaryeng 11y ago
I don't think you know how these work
- ScottBurson 11y ago
  tosseraccount has an excellent point. Overfitting is a huge danger. In the end only trading the live market for some time will tell you whether you have a profitable strategy.

brobinson 11y ago

Your algorithm will likely not stand up to Walk Forward Analysis/Optimization/Testing: https://en.wikipedia.org/wiki/Walk_forward_optimization https://en.wikipedia.org/wiki/Walk_forward_optimization I recommend reading "The Evaluation and Optimization of Trading Strategies" by Robert Pardo.

solve 11y ago
It's really the most fundamental rule for every type of machine learning task -- there should be both a (1) test set and (2) a separate holdout validation set. No matter if the task is classification, sequence prediction, regression, manifold learning, clustering, whatever. Really surprising how so few amateur traders understand this fundamental principal. Scientists too.

MaxScheiber 11y ago

> Buy and hold forever is generally great advice, until it isn't. That's a truism. I could say the same thing about algorithmic trading. The fact of the matter is that the median hedge fund manager has negative alpha. It is extremely difficult to pick winners via any strategy. In fact, it's so difficult that for people who can do it, the resulting lifestyle and compensation is so great that the asset management shops that they work out have almost zero attrition. Moreover, the whole point of passive investment is that the odds of that kind of a market-wide crash affecting your money over a 30-year timespan are infinitesimal. If the market actually were to devolve like that, nobody would make money in the market, regardless of strategy. > Investors should be looking for ways to systematically protect themselves from the market's irrational mood swings. Yes, it's called diversification. I won't harp on the backtesting point, but what I will say is this: I'm sure that you understand the risk-reward tradeoff. It would be impossible to participate in any sort of finance and not know this concept. Quite frankly, the probability of 30%+ of your life savings being wiped out in an index fund is unbelievably low. You eat some return at the expense of lower volatility, but we're talking about a 401(k) here. We want that kind of expected stability.

tosseraccount 11y ago
"the probability of 30%+ of your life savings being wiped out in an index fund is unbelievably low" The crash of 69 : 36% The crash of 73: 48% The crash of 87: 33% The dotcom crash 2000: 50% Housing Bubble 2007: 56% [ source : http://traderhq.com/illustrated-history-every-s-p-500-bear-market/ http://traderhq.com/illustrated-history-every-s-p-500-bear-m... ] The odds that at some point a portfolio of S&P 500 stocks will be 30% off a recent high is actually very high. Wars. Panics. Depressions. Plagues. Recessions. Revolutions. They still happen.
- jordan0day 11y ago
  I don't know much about the market, and know even less about actual investing strategy-ish stuff, but what does the loss look like five years later? I guess what I mean is, for an "index fund", crashes don't seem to be... permanent? Perhaps that's small comfort for someone who's set to retire the month after one of these 30% crashes, but I guess that's what the whole "invest more conservatively as you get older" thing is supposed to help with?

nicholas73 11y ago

The only quibble I have is that taxes would outweigh the advantages here. You either do this in a tax free account, or your trading needs to be a lot better!