Improving the Timing of Stock Purchases
Issue: It is possible for mere mortals to improve the timing of stock purchases
Premise: It is human nature to allow emotional issues to effect decisions. The words Fear and Greed are commonly heard on Wall St: Fear causes us to sell at the wrong time (I had better sell now, even though the price is low because the market is down and it may fall further) while Greed causes us to buy at the wrong time (I feel good, the market is climbing, my stock is soaring, I had better buy now before I miss it). Many people feel the urge to buy when the market is having an up day and to sell on down days to protect themselves, but with roller coaster markets this can lead to disaster. There must be a better way. (Note: Frank's buys were placed on both up days (50%)and down days (50%)on the Nas and sometimes QCOM was up and sometimes it was down.)
Idea 1: Buy without regards to market timing. Many very smart people believe that it is impossible to time the market, and they are probably right. (Although, as an aside, timers seem to fare better in down trending markets such as post 1989 Japan or the Gold market for almost 20 years.)
Idea 2: Can we perhaps use a mechanical system to improve our buy points.
Important Note: This is just an idea that is being back-tested here on just six buys. Real tough to draw conclusions on such a small, unscientific sample, but lets do it anyway.
The Data: Uncle Frank posted his last seven QCOM buys. However, there is a bit of confusion on the last one, so I am just going to consider the first six. Here they are:
4/5 @ 70 3/8
4/13 @ 84 3/4
5/6 @ 111 1/4
5/13 @ 116 3/8
5/21 @ 100 5/8
6/21 @ 127 «
The system: Instead of placing a buy order, place a good until canceled order at a price 10% below the buy price. Keep the order in place for three days after the buy date, if near the end of the third day the order has not gone through, cancel it and buy the stock at the market price.
Hypothetical Results: The first GTC order goes in for 63 3/8, but it does not go off, forcing us to buy the stock at 74 1/4 on 4/8
The next GTC order goes in at 76 1/4 and the buy goes through on 4/15.
The third GTC is for 100 1/8. It is not triggered, but we still get the stock for a lower price at 109 «
The fourth GTC is for 104 3/4 which is executed two days later on 5/17
The fifth GTC is for 90 « and it too is executed two days later
The sixth GTC is for 114 3/4 but it is not executed and we end up buying on the third day for 127 3/4.
The Result: Out of six tries we got the stock at a price 10% lower than our original price in three cases (50%). In two cases we got the stock for about the same price, while in one case, the first, we ended up paying quite a bit more - which in real life might have caused us to toss the system out the window. And, a word of warning, there is always the risk that something dramatic could happen and the stock could vault much high causing a buy at a much higher price.
The average price of Frank's six buys was about 101 3/4 while the average of the six using "the system" was about 97 1/8. If 100 share lots were purchased each time the system would have saved $2800.
Why 10%: My first idea was to place each GTC at $5 less than the buy prices and that worked, (in fact the first buy would have been executed at 65 3/8) the average being about 97 3/4. So then I thought, lets be greedy and try $10 below and that worked too, with an average of about 97 1/4. But then I thought since $5 or $10 is so arbitrary a new number would be needed for each situation depending on the stock price of the issue and its volatility - I mean it would be silly to place GTC orders $10 under the price of a $20 stock and expect them to go off. So I came up with a percent (on the high side) that could be used in any situation.
Will this work in real time? Who knows, but it might be worth thinking about.
StockHawk |
