I also am having a hard time understanding why you’re attaching such significance to the opening price. If a stock is trading higher or lower than its opening print that doesn’t mean it’s trading at a premium/discount to its “true” value.
It doesn’t have to be monotonically increasing. Look, the price of an given stock is essentially a random walk with an upward trend. Since there’s an upward trend, you’re better off buying as soon as possible. Waiting for a particular pattern to show up in the random walk doesn’t improve your performance any. In fact, it decreases it, because you’re waiting.
What you still seem not to understand is that there’s nothing special about the opening price. Waiting for an intra-day fall in price won’t make you get a discount; it’ll just make you wait (which is on average bad, because the price goes up over time). It seems like a good idea because it seems kind of like a store putting something on sale, but that’s not how stock market prices work. Basing your purchase decision on recent small changes in the price is just as arbitrary deciding to buy stock only in the three days after a full moon.
But don’t take my word for it. Seriously: Pick a stock at random and a few hundred dates at random and test your plan against the “just buy it immediately plan”.
Sorry, made an error. Test results to be posted shortly.
Ok, I got curious and ran some numbers on SPY (the S&P500 ETF). I downloaded data between 1993 and 2009 from Yahoo, and ran the following algorithm on it, testing every day:
If the low price for the day is below the opening price (by some threshold), buy one share on that day. Otherwise, buy that share on the next day for which that’s true. This keeps the number of shares purchased the same for all runs. At the end of the test (on the “last” day), there may be some shares left unpurchased. I’m not sure what to do with these, so I just output the count. The first one is with a threshold of 0, so it just buys every day. The last one is clearly not doing a lot of buying. 10% single-day drops aren’t very common.
Here are my results:
Strategy of waiting for a 0.000000 drop cost an average of 102.022586 per share. 0 days left at end
Strategy of waiting for a 0.001000 drop cost an average of 102.010184 per share. 0 days left at end
Strategy of waiting for a 0.002000 drop cost an average of 101.983278 per share. 1 days left at end
Strategy of waiting for a 0.003000 drop cost an average of 101.977852 per share. 1 days left at end
Strategy of waiting for a 0.004000 drop cost an average of 101.964396 per share. 1 days left at end
Strategy of waiting for a 0.005000 drop cost an average of 101.958242 per share. 1 days left at end
Strategy of waiting for a 0.006000 drop cost an average of 101.954964 per share. 1 days left at end
Strategy of waiting for a 0.007000 drop cost an average of 101.962944 per share. 1 days left at end
Strategy of waiting for a 0.008000 drop cost an average of 101.978187 per share. 1 days left at end
Strategy of waiting for a 0.009000 drop cost an average of 102.042020 per share. 1 days left at end
Strategy of waiting for a 0.010000 drop cost an average of 102.076804 per share. 1 days left at end
Strategy of waiting for a 0.020000 drop cost an average of 104.379023 per share. 7 days left at end
Strategy of waiting for a 0.030000 drop cost an average of 106.541088 per share. 81 days left at end
Strategy of waiting for a 0.050000 drop cost an average of 103.696891 per share. 101 days left at end
Strategy of waiting for a 0.100000 drop cost an average of 0.000000 per share. 4137 days left at end
So, waiting for a small (< 1%) drop appears to show a slight (very slight: < 0.1% over the naive “just buy” dollar-cost-average strategy) improvement, at least in this test data. Waiting for a larger drop is clearly worse. I believe that the apparent slight improvement is just an artifact of the test data, but I don’t have time to run more tests right now. Maybe tomorrow.
ETA: Just realized that this isn’t taking into account dividends. Not sure if dividends cause intra-day drops, so I don’t know if this is an issue for my model.
If you’re using Yahoo Finance’s historical prices, the Adj Close column is adjusted for dividends and stock splits.
Thanks walrus, those numbers are very interesting.
To answer the question about why use the opening price, my answer is why not? The trading company I use lets you set a limit order that expires that day or in 60 days. So I chose “opening price” as a starting point. If you can show a different metric produces different results, I’m all ears.
Hmm. I see that. I’m not sure how I’d incorporate that into my model, since I don’t use the closing price for anything.
OK. So you agree that the opening price is completely arbitrary and no more “real” than any other price throughout the day. The use of the word “discount” is what threw me.
Ahh, sorry, you’re right.
When a stock/ETF pays a dividend, the price does indeed drop by the amount of the dividend. This happens overnight on what is known as ex-dividend day. As an example, if a stock closes at $10.00 the day before it pays a $0.50 dividend, it will open at $9.50, ceterus paribus, the next day.
Do you have a PHD in statistics and a business degree or other extensive experience in financial market dynamics? Have you listened to finance professors lecture on market efficiency and present their investment strategies research that they’ve back-tested for every stock for every hour of every day for the last 80 years (or as far back as there’s data)? Have you performed your own analysis on this level?If the answer is no you should probably run away very quickly (just like me and 99.99% of the world population). There’s a lot of other super smart people who will eat your lunch as you try to beat the market. (And I’m not saying they’re perfect either, but that just illustrates the difficulty of beating the market).
Discussing the merits of different strategies is fun, but saying they are correct without adequate experience and running reams of test data is akin to saying “I’m going to be a NFL running back and make a lot of money by just not getting tackled. It’s so obvious, why isn’t everyone doing it, I just must be smarter than everyone.”
At the end of day, any stock that you can routinely buy at X% discount de facto has a lot of intrinsic volatility that will bite you in the butt at some point and wipe out everything you thought you’d won. 99.99% of investors can’t beat the risk/reward curve of the market. They just confuse higher reward with beating it without understanding that the risk is higher still and may be delayed a bit.
I disagree.
My stock strategy over the years could probably be summarized as:
“Buy High, Sell Higher”
What your strategy will likely do is work for most and make peanuts only to be wiped out with stocks getting a huge downswing.
Keep in mind if a stock gets hammered…lets say real bad news and gets hit for a 20% loss…YOU will now own that stock with a 18% loss. Not a good day.
I also have gotten away from limit/stop orders. I know, I know it is the conventional ‘wisdom’ to use them and if you don’t you are a fool etc etc etc. However, I have seemed to notice that these orders seem to trigger at the times most disadvantagous to me…it’s like people KNOW* what my order is and can manipulate the stock to trigger it.
*I found out that they DO know…there are services out there that allow you to see the limit/stop orders for the stock. I know this is true because I checked it out. On thinly traded stocks even I could muster the resources to snag a juice stop order. I have no doubt now that that happened to me.
The dollar cost average strategy works because it smooths out the fluctuations in the share price over time. Sometimes you buy low and sometimes you buy high. But as long as the general trend is upward and the stock is fundamentally sound, you will make money in the long run.
Your flaw with C is that you seem to assume that the stock price automatically fluctuates from the opening price in a predictable manner. i.e., it dips and then automatically rises.
I see no mention of what price you think you should sell it at.
Selling the stock is currently outside the scope of discussion. Why is that so interesting to you?
If your method worked, then many people would already be using it, such that the return from your method would become the new index, and you’d be back to not beating the index.
Long ago I saw a book based on a similar strategy. (IIRC, it was called The Money Grinder, from Self-Counsel Press in Vancouver, but now it isn’t listed on their web site. I didn’t buy it then, and wouldn’t now anyway.)
The idea was to set up an account with equal amounts of cash and stock, calculate low and high price limits according to the author’s formulas, and buy or sell a block of your stock whenever a limit is reached to keep the cash and stock amounts equal. Then recalculate and proceed.
Of course if the stock keeps going up you sell out of stock early, and if it keeps going down you spend all your cash. The author didn’t seem to consider either of those possibilities.
For one thing, we won’t know if your strategy works without knowing how many losses you take on the losers and how much you make on the winners.
Let’s say your strategy produces a string of 10 +1% gains, but you set no provision for selling the stock when it goes down, so that’s capped off by 1 -100% loss. Congratulations, your plan has netted you -90% returns. (And, yes, clearly an index stock is going to lose something less than 100%. But I hope you haven’t forgotten a year not so long ago in which they lost 50%.)
Alternatively, let’s say your strategy involves no selling whatsoever. That means that you may start with a 1% advantage over Joe, who bought at the opening. However, over 20 years neither of you sell. So now you bought $99 of stock now worth $1,000 and he bought $100 of stock now worth $1,000. Your masterful manipulation of the stock market will have netted you an extra 0.1% over Joe.
You don’t need to discuss a selling strategy to determine whether the buying strategy will save you money. All you need to do is figure out how much you’d spend on buying the stock in different scenarios where you end up with the same stock.
Yeah, I can see that in a general sense. It just seems like waiting for a price drop has implications about the direction an index/market is already heading. A buying strategy that has you jumping into the market only as 1929 and 2007 start crashing is a guaranteed failure for short-term trading because it took some time for markets to bottom out. What looked like a 10% “discount” in the early days of the crash would turn into a 40% loss at the bottom. You’d be holding for decades just to recoup the losses.
Nobody is fooled by your skewed narrative. For the exact same scenario, I would say the protagonist buys an extra 1% of stock and after 20 years, it’s ten times the difference you calculated.
But whether it’s 1% or 0.1%, it’s still extra money for you. According to a Capital One commercial, research indicates everyone wants more cash.
Because you don’t actually make any money on a stock until after you sell it.
And as others have already pointed out, you aren’t purchasing the stock “cheaper than a regular trade”. You purchasing the stock at a market price you think will occur some time after the opening bell.