1nvest wrote:AFAIK AIM wont ever sell out of shares, but can exhaust cash reserves. There's some detail of the method here
https://www.gummystuff.org/AIM.htmTends to be a cash-cow, throws off lots of cash, which is useful for those in drawdown. I apply it to the S&P500 index/price once/month adjusted for inflation, which is a bit speculative i.e. a approximation of the most recent (US) inflation CPI index each month ahead of actual formal reporting. Periodically I revise the history in reflection of actual reported inflation rates and find that the differences tend to be relatively small. Standard AIM is applied to nominal price, I find that applying it real price worked better, otherwise I use the exact same method, 10% SAFE, 5% of stock value minimum trade size, monthly reviews.
I just use it as a guide, all on paper. It tends to do OK at indicating appropriate levels of risk to hold over time so is a form of comfort when buying into dips when otherwise fear of further declines to come might have you failing to actually trade/buy. RIP Robert Lichello is saying its appropriate to trade now, to x% amounts and he's more often been right in the past type comfort.
I have long been intrigued by automatic trading strategies, such as AIM, but have never actually run one for any length of time, although I have investigated several home-brew algorithms by back testing. I came to the conclusion that, in certain periods, they could perform “quite well”, but it is very easy to be deceived about an algorithm’s behaviour by ignoring results that turn out bad, or not even knowing about ones that might have turned out well - both very human tendencies. In particular, the AIM method, which I have not tried in the form suggested by your link above, has features which common sense suggests that any strategy should have, i.e.
- tries to stay mostly invested
- a cash buffer
- limits on transaction size and frequency
- buy low, sell high
and some which leave me wondering about it, i.e.
- why 10%? Maybe 12% or 8% would work better. After all, we are always being told that small increments over time, such as 1.5% fund management charges, can make a big difference.
- why is the cash buffer the size that it is and why isn’t its size included in the decision to buy or sell?
- The PC (Portfolio Control) only ever goes up. Why? In a long bear market, this may not be desirable.
- why not reinvest dividends?
- if we diversify our stocks, should they share a common cash buffer?
But I have not read the book, or played with the algorithm myself, so I may be misunderstanding the examples I found on the Internet, e.g. here
https://toughnickel.com/personal-finance/robertlichelloAIMSystemAnd another thing, back testing which compares LTBH with a strategy such as AIM needs to be done on a like-for-like basis. The only way to compare the two is to run an IRR calculation based on the timings of the buy, sell and dividend cash flows. This has an important bearing on risk, because the percentage of the portfolio in cash at any moment determines the risk exposure (assuming that holding cash has zero risk, which it doesn’t). Some of the algorithm tests that I have tried resulted in lower total P/L but higher IRR than LTBH with the same starting amount invested. This can happen when holdings are consistently reduced when the stock price is high and increased when it is low, resulting in a better return/risk profile.
This has got me thinking,
regards,
S