The normal protocol for holding stock is, to hold stock. It's a good protocol. (The best strategy I know is also the simplest: put 100% in QQQ, hold it until you need to spend money.)

Here's another protocol for holding stock:

• When the stock price goes up 56.25% from your last buy/sell, sell 20% of your current holding, buy QQQ instead.
• When the stock price goes down 36% from your last buy/sell, buy 25% more than your current holding, get the cash by selling QQQ.
• If you need cash to spend on (for example) taxes or household expenses or other stocks, sell QQQ.

This protocol has a couple properties:

• If the stock drops to 0, you end up losing 2x your original purchase. That's not as good as the 1x loss if you had just held, but at least it's not infinite.
• When the stock goes up 56.25%, even after selling 20%, your holding is still worth 25% more than at your previous buy/sell. That's not as good as the 56.25% as if you had just held, but at least it goes up. Plus you get 31.25% of the original purchase price worth of QQQ.
• If the stock goes up then down then you make money. If it starts at X, goes up 56.25%, then back down to X, and QQQ doesn't move at all, you have the same amount of stock you started with, yet you also have 11.25% of your the original worth in QQQ.
• If the stock goes down then up then you make money. If it starts at X, goes down 36%, then back up to X, but QQQ doesn't move at all, you have the same amount of stock you started with, yet you also have 9% of your original worth in QQQ.
• This gives you an excuse to watch the stock market and makes you feel like you're doing something useful when stocks go up or down, which may prevent you from deciding to do stupid things, like going all-cash at the bottom of stock market crashes.
• You could use cash instead of QQQ, but QQQ goes up 9% per year on average while cash goes up 0%, so I find QQQ a better default than cash. If you need spending money, you usually don't have to decide which stock to sell, the default is sell QQQ. If you want to add money to your account, you don't have to decide what to invest in right then, the default is to buy QQQ. Ideally the 56.25% and 36% ratios would be between the stock price and QQQ price rather than just the absolute stock price, but I always do the absolute stock price because it's easier.
• 56.25%, -20%, 36%, -25%, and 2x are really pp-1, 1-1/p, 1-1/pp, p-1, and 1+(1-1/p)/(1-1/p) where p=5/4. Some other values of p:

  p ratio	sell price	sell amount	buy price	buy amount	up-down gain	drop-to-zero loss
  20/19	10.8033%	-5%	-9.75%	5.2632%	0.5402%	2x
  10/9	23.4568%	-10%	-19%	11.1111%	2.3457%	2x
  5/4	56.25%	-20%	-36%	25%	11.25%	2x
  3/2	125%	-33.3333%	-55.5555%	50%	41.6667%	2x
  2/1	300%	-50%	-75%	100%	150%	2x

As near as I can tell this protocol is worse than just holding stock. Because stocks don't wiggle enough compared to how often they go just up or just down. If you compare whether p or pp is better, for any p for most stocks over most time ranges, pp is better, and the limit of that (infinite p) is just holding stocks. This protocol doesn't say how to decide what to invest in in the first place, it is just a protocol for holding stock that you've already decided to invest in.

If p=1+d, the gain of that strategy from an up/down grows with the square of d. You gain 4x as much from an up/down with 1+2d than you would with 1+d. So 1+d and 1+2d would be equally good if it was 4x more likely to cross buy/sell points with 1+d than with 1+2d. It's actually about 3x more likely, and that is fairly consistent all the time across many stocks. For any 1+d it makes money, just not as much as buy and hold.

A complementary strategy bets that up-down is less likely than up-up or down-down. If you see a down-up, you would buy, on expectation that it will continue up. As long as it goes up you would hold. If you see a up-down, you would sell, on the expectation it will continue down. As long as it goes down you'd ignore it. This also tends to make money, but not as much as buy and hold.

Since the two methods both make money and are based on the same signals, you can add them together, and this gives you the sum of their profits. Repeated up-down-up-down has the two strategies do opposite things, so they cancel out. The sum has you buy on each down-down and sell on each up-up. For example, 1up-2down-3down-4down-5down-6up-7down-8up-9up would be 1nothing-2nothing-3buy-4buy-5buy-6nothing-7nothing-8nothing-9sell.

This combined method is still worse than buy-and-hold.

The strategies above doesn't say which stocks to play in the first place. Right now I'm choosing based on a combination of rising revenues and not-terribly-high price/sales ratios, and selling when revenues consistently shrink. But that strategy has not served me well. I got a lot of medium price/sales ratio stocks with high growth which then became low price/sales ratio stocks with little to no growth. And it's hard to distinguish between longterm dropping revenues and dropping revenues due to shortterm unusual circumstances.

If you look at long-term stock trends, it looks like a bouncing ball: rounded peaks and sharp bottoms. I've had any number of times where I sold a stock and it doubled within the next two weeks. So now, when I decide it's time to sell a stock, I make the decision but wait two weeks after that before actually selling, and only if it is still worth selling. That seemed to fix the issue. Buys I also wait, but only two days.

Bob Predicts the Future
Table of Contents