5 ms·

"you may as well be doubling your bet every time you lose at roulette" - in trading it's called DCA or Dollar Cost Averaging. Works ok/good for very long term i

"you may as well be doubling your bet every time you lose at roulette" - in trading it's called DCA or Dollar Cost Averaging. Works ok/good for very long term investments in stable assets like SP500 Index Funds. But for daytrading the rule of thumb is "don't add to your losers". There's a good reason for that.

hansvm 6y ago
DCA isn't a panacea. It performs well in some scenarios and poorly in others (e.g. if by some means you're somehow certain the market will have a large dip and recovery but you don't know anything about the timing then dollar cost averaging will outperform many alternatives). Applying a rule of thumb as simple as "for long term investments in stable assets" misses enough nuance to lead people astray.
- jiofih 6y ago
  It literally has 100% guaranteed return over a period of 20 years, going back all the way to the first stock exchange. Not the best returns though.
  - hansvm 6y ago
    It's easy to pick a winning strategy when everything is going up. If that's the assumption you're using for your investments though (everything has been going up, so it'll keep going up) then DCA still probably isn't what you want -- it'll underperform even dead simple strategies like a diversified buy and hold nearly always. My point is that when choosing investments if you have any theory at all as to what the market might do (and if you don't, why are you investing?), you ought to choose a strategy which based on that information brings you close to your goals, and secondarily it's worth pointing out that in a broad variety of pretty ordinary market conditions DCA is not a good choice.
    - jiofih 6y ago
      Exactly. That theory is that the market will always go up long term (because capitalism is inherently inflationary) and historically, DCA plus holding has always given returns over a > 20y period with any reasonably mixed basket. A “diversified buy & hold” assumes you have your entire 20y investment capital available at the outset, which is rarely the case.
      - hansvm 6y ago
        > A “diversified buy & hold” assumes you have your entire 20y investment capital available at the outset, which is rarely the case. If you don't then you're effectively implementing DCA because you don't have better alternatives available -- regardless of the strategy you use, if you immediately invest spare funds as soon as they're available from a source of evenly distributed recurring income then it'll look a lot like DCA. That's still consistent with my position of using the data you have to pick the best option you can. If you have options other than DCA (e.g., a sizable percentage people get some kind of windfall in their life), then carefully evaluate whether DCA is the right way to treat that capital. It usually won't be, even compared with dead simple strategies like a diversified buy-and-hold. As an aside: Ignoring any kind of extreme luck, in a field like tech with rapid raises, no matter which investment strategy you choose your nest egg will almost entirely be comprised of funds from your last 5-10yrs of work.
  - RussianCow 6y ago
    Nothing is "100% guaranteed". There are lots of reasons that the returns of investing in index funds using DCA may dry up, or even fall over the long run—for instance, it's not inconceivable that a significant enough market downturn could bankrupt a lot of Fortune 500 companies, bringing the indexes down significantly for a long period of time. Even though these types of scenarios seem extremely unlikely, it's incredibly inaccurate to say that any such returns are "100% guaranteed".
    - jiofih 6y ago
      Yes, you can argue that, but it’s statistics, and includes the three big crashes the market has lived through. It’s a 20 year period for a reason - it’s how long it would have taken you to recover from 1929 if you bought at the peak just before the crash.
      - RussianCow 6y ago
        > It’s a 20 year period for a reason - it’s how long it would have taken you to recover from 1929 if you bought at the peak just before the crash. You've cherry-picked an arbitrary window based on historical returns, but there is no reason that this will hold in the future. You can come up with lots of strategies with arbitrary conditions that would have given you great returns throughout the past century, but completely fail after you implement them. I don't see why dollar cost averaging is any different. > it’s statistics Exactly. And nothing in statistics is guaranteed. The fact that it's worked well 100% of the time in the past does not imply that it will work 100% of the time in the future. Does it seem highly unlikely that DCA will stop being an effective investment strategy in the foreseeable future? Yes. But I take issue with the "100% guaranteed" claim, and everyone should realize that no investment strategy is completely risk-free.
        jiofih 6y ago
        Well, down here in the real world, a 100% statistical chance is as good as it gets. Why would you try any other strategy with historical success < 100%, or more likely, unknown?
- deandree 6y ago
  Nothing is 100% and/or guaranteed in trading. It's a game of probabilities and risk management
  - hansvm 6y ago
    Sure, but investing on a rule of thumb is neither an exercise in probability nor risk management. I'm advocating for using the information you have to choose an appropriate strategy, and I only brought it up at all because of how many people I've seen using DCA inappropriately because they don't understand when it performs well and when it doesn't.

hansvm 6y ago

DCA isn't a panacea. It performs well in some scenarios and poorly in others (e.g. if by some means you're somehow certain the market will have a large dip and recovery but you don't know anything about the timing then dollar cost averaging will outperform many alternatives). Applying a rule of thumb as simple as "for long term investments in stable assets" misses enough nuance to lead people astray.

jiofih 6y ago
It literally has 100% guaranteed return over a period of 20 years, going back all the way to the first stock exchange. Not the best returns though.
- hansvm 6y ago
  It's easy to pick a winning strategy when everything is going up. If that's the assumption you're using for your investments though (everything has been going up, so it'll keep going up) then DCA still probably isn't what you want -- it'll underperform even dead simple strategies like a diversified buy and hold nearly always. My point is that when choosing investments if you have any theory at all as to what the market might do (and if you don't, why are you investing?), you ought to choose a strategy which based on that information brings you close to your goals, and secondarily it's worth pointing out that in a broad variety of pretty ordinary market conditions DCA is not a good choice.
  - jiofih 6y ago
    Exactly. That theory is that the market will always go up long term (because capitalism is inherently inflationary) and historically, DCA plus holding has always given returns over a > 20y period with any reasonably mixed basket. A “diversified buy & hold” assumes you have your entire 20y investment capital available at the outset, which is rarely the case.
    - hansvm 6y ago
      > A “diversified buy & hold” assumes you have your entire 20y investment capital available at the outset, which is rarely the case. If you don't then you're effectively implementing DCA because you don't have better alternatives available -- regardless of the strategy you use, if you immediately invest spare funds as soon as they're available from a source of evenly distributed recurring income then it'll look a lot like DCA. That's still consistent with my position of using the data you have to pick the best option you can. If you have options other than DCA (e.g., a sizable percentage people get some kind of windfall in their life), then carefully evaluate whether DCA is the right way to treat that capital. It usually won't be, even compared with dead simple strategies like a diversified buy-and-hold. As an aside: Ignoring any kind of extreme luck, in a field like tech with rapid raises, no matter which investment strategy you choose your nest egg will almost entirely be comprised of funds from your last 5-10yrs of work.
- RussianCow 6y ago
  Nothing is "100% guaranteed". There are lots of reasons that the returns of investing in index funds using DCA may dry up, or even fall over the long run—for instance, it's not inconceivable that a significant enough market downturn could bankrupt a lot of Fortune 500 companies, bringing the indexes down significantly for a long period of time. Even though these types of scenarios seem extremely unlikely, it's incredibly inaccurate to say that any such returns are "100% guaranteed".
  - jiofih 6y ago
    Yes, you can argue that, but it’s statistics, and includes the three big crashes the market has lived through. It’s a 20 year period for a reason - it’s how long it would have taken you to recover from 1929 if you bought at the peak just before the crash.
    - RussianCow 6y ago
      > It’s a 20 year period for a reason - it’s how long it would have taken you to recover from 1929 if you bought at the peak just before the crash. You've cherry-picked an arbitrary window based on historical returns, but there is no reason that this will hold in the future. You can come up with lots of strategies with arbitrary conditions that would have given you great returns throughout the past century, but completely fail after you implement them. I don't see why dollar cost averaging is any different. > it’s statistics Exactly. And nothing in statistics is guaranteed. The fact that it's worked well 100% of the time in the past does not imply that it will work 100% of the time in the future. Does it seem highly unlikely that DCA will stop being an effective investment strategy in the foreseeable future? Yes. But I take issue with the "100% guaranteed" claim, and everyone should realize that no investment strategy is completely risk-free.
      - jiofih 6y ago
        Well, down here in the real world, a 100% statistical chance is as good as it gets. Why would you try any other strategy with historical success < 100%, or more likely, unknown?
deandree 6y ago
Nothing is 100% and/or guaranteed in trading. It's a game of probabilities and risk management
- hansvm 6y ago
  Sure, but investing on a rule of thumb is neither an exercise in probability nor risk management. I'm advocating for using the information you have to choose an appropriate strategy, and I only brought it up at all because of how many people I've seen using DCA inappropriately because they don't understand when it performs well and when it doesn't.