As of 09/15/2021
Indus: 34,814 +236.82 +0.7%
Trans: 14,420 +134.87 +0.9%
Utils: 925 +0.27 +0.0%
Nasdaq: 15,162 +123.77 +0.8%
S&P 500: 4,481 +37.65 +0.8%

YTD
+13.7%
+15.3%
+7.0%
+17.6%
+19.3%

35,600 or 33,800 by 10/01/2021
15,400 or 14,000 by 10/01/2021
965 or 905 by 10/01/2021
15,600 or 14,700 by 10/01/2021
4,650 or 4,350 by 10/01/2021

As of 09/15/2021
Indus: 34,814 +236.82 +0.7%
Trans: 14,420 +134.87 +0.9%
Utils: 925 +0.27 +0.0%
Nasdaq: 15,162 +123.77 +0.8%
S&P 500: 4,481 +37.65 +0.8%

YTD
+13.7%
+15.3%
+7.0%
+17.6%
+19.3%
 
35,600 or 33,800 by 10/01/2021
15,400 or 14,000 by 10/01/2021
965 or 905 by 10/01/2021
15,600 or 14,700 by 10/01/2021
4,650 or 4,350 by 10/01/2021
 
Initial release 5/25/2021.
I was considering buying more Beazer stock (BZH, see chart) but it was up 51% yeartodate. How much higher can I expect it to rise? To answer that, I did some research.
I started with stocks priced above $5 a share (because those priced lower, often by stock splits, showed minor price fluctuations which translated into large percentage moves). I found the yearly low (the lowest price for the year) followed by the highest price within a year, after the lowest low. If price continued to make new highs after the year, then I let the move continue until the stock dropped from a high to close at least 20% lower.
Between the lowest low and highest high, I counted the number of times price dropped at least 20% along the way. A 20% drop from high to close is what I call a trend change. You'd want to sell before one of those happened. A 20% decline from a high is also what determines the start of a bear market in the indices, so it's a good benchmark.
I used 468 stocks from February 1981 to May 2021 and found 8,601 samples from bull markets (but I didn't exclude the Covid19 drop in 2020).
Refer to the figure on the right. Here's what I found from the data.
Let's begin with how far price rises from the yearly low to the following 1year high. That's the highest high one year after the yearly low and continuing higher until price dropped at least 20%, measured from high to close. It's the AH rise shown in the prior chart.
Gain  25%  50%  75%  100%  125%  150%  175%  200%  225%  250%  >250% 
Frequency  8%  20%  20%  15%  10%  7%  5%  4%  3%  2%  8% 
Cumulative  8%  28%  48%  62%  72%  79%  84%  88%  90%  92%  100% 
Returning to my puzzle with Beazer stock and its 51% gain, the table suggests there's more upside to come with just 28% of the stocks failing to rise more than 50%. That leaves the other 72% of stocks which saw price climb higher.
Having said that, the Beazer stock measure was yeartodate (chart at top of page), and not taken from the yearly low. Going back to the full year of 2020 (chart on the right, weekly scale), the stock hit a low of 4.39 in early 2020 and 14 months later, it reached a high of 26.12. That's a 595% move, shooting it off the right end of the table. (Because the stock dropped below $5 a share, this uptrend wasn't included in my research of stock rises).
Along the way, the stock showed a 20% decline starting on 26 March 2020 (a week after the yearly low) and another on 8 June 2020 (see chart), although it's hard to see them on the weekly scale.
The duration of the uptrend is 14 months (so far), shown from the yearly low to the red vertical line (1 year after the low) and extending to the right of the line.
Considering that stock is up 50% YTD coupled with that huge 595% move, the end of the run is probably close to being over. The housing market is overheating in my opinion, so I'm watching it for weakness and getting ready to sell my position.
Next, let's examine each leg to see how far price is likely to climb over the course of a year. The average rise was 110% (A to H in the chart). Just over half the stocks (54%) had an uptrend which lasted beyond the year. Recall, I found the highest high after the lowest low within a year but continued the uptrend search until the stock dropped at least 20%, measured from high to close, regardless of how long it took to get there. The Beazer chart is an example of this extended move higher.
Of the 8,601 samples I looked at, just 28% of them had a first leg which saw price drop at least 20% along the way (and recovered to a new high) from A to H. On the Results chart, I show that as the BC drop.
The first leg rise is AB. It ends with a 20% hightoclose drop in the stock (B to C). The average rise from A to B measured 69% with a 58% median. Here's a frequency distribution of the rises in the first leg.
Gain  30%  40%  50%  60%  70%  80%  90%  100%  125%  150%  >150% 
Frequency  2%  17%  19%  15%  12%  9%  6%  5%  7%  4%  4% 
Cumulative  2%  19%  38%  53%  65%  74%  80%  86%  92%  96%  100% 
As the table shows, half the samples (53%) saw price rise less than 60%. So if you have a stock which has climbed more than 60%, then you're likely closer to the end of the trend than the beginning. Just 14% of the stocks will see price rise more than 100%.
The second leg measures the rise from peak B to D followed by a drop of 20% (or more) from D to E. Here's a frequency distribution of the BD move. Just 551 samples had a second leg.
Gain  5%  10%  15%  20%  25%  30%  35%  40%  45%  50%  >50% 
Frequency  11%  18%  15%  10%  9%  7%  6%  3%  3%  3%  14% 
Cumulative  11%  28%  44%  54%  63%  70%  77%  80%  83%  86%  100% 
Notice that the scale has changed from increments of 10% to 5%. Just over half (54%) of the 551 samples will see price rise 20% in this leg (the median is 18%).
Twentythree percent of the samples having a first leg will see a second leg. In other words, 23% will see price drop at least 20% twice on the journey from A to H. Also, price in the second leg will rise only a third as far as they did in the first leg.
If you calculate the first leg rise, take a third of that and add the result to the prior peak (before the 20+% drop), you can have a price target.
The chart shows the third leg as the move measured from peak D to F with a 20% or more decline to G. Only 153 samples made the cut this time, or 28% of those which had second legs will form a third leg on the rise to the oneyear high. Here are the numbers.
For targeting, use the height of the second leg rise as the proxy of what to expect during the third leg.
Gain  5%  10%  15%  20%  25%  30%  35%  40%  45%  50%  >50% 
Frequency  10%  13%  13%  12%  12%  5%  8%  5%  3%  2%  17% 
Cumulative  10%  23%  36%  48%  60%  65%  73%  78%  81%  83%  100% 
The median rise for this leg is 21%, measure from the top of the second leg (D) to the peak at F.
Just 44 samples had a fourth leg which ended with price dropping at least 20%. I show that in the above chart as the rise from peak F to H with a drop to I.
Gain  5%  10%  15%  20%  25%  30%  35%  40%  45%  50%  >50% 
Frequency  5%  14%  9%  16%  11%  7%  11%  0%  5%  5%  18% 
Cumulative  5%  18%  27%  43%  55%  61%  73%  73%  77%  82%  100% 
The median of this rise was 23%, meaning half the samples showed a smaller rise and half had a bigger one.
For targeting, you can use the height of the third leg as the proxy for the anticipated height of the fourth leg.
For trading, I told you about my Beazer holding. I decided not to buy more of the stock, but to hold onto what I already owned and watch it carefully. Instead, I bought shares of an ETF (exchange traded fund) to avoid single stock risk. That risk happens when a stock doesn't deliver the kind of earnings performance that the market expects. The stock can drop up to 60% or more, at the open. That's rare but it does happen, especially when price climbs to the upper reaches of the atmosphere.
The ETF I bought is a blend of housing stocks, allowing me to participate in the potential continued upside rise while avoiding single stock risk.
Let's take an example and see how we can use the findings in this article to predict how far price might rise.
I show a chart of 3D Systems (DDD) on the weekly scale. C is a symmetrical triangle highlighted by the converging red lines. The bottom line shows where trendline resistance is expected. However, the apex of the triangle I believe is a more potent stopping place. That looks to be about 16 on this chart (eyeballing it. Later we'll see that turn F was at 15.94, just pennies away from the apex).
Point B is the 2017 yearly low. Point A is the first week of the new year. Recall that I begin the search for the start of an uptrend by using the yearly low. By looking at this chart, it's easy to see the 2017 yearly low but we have no idea where it will be in 2018. However, we are given a clue.
The apex of the triangle, D, is almost directly above A. The apex of a triangle is a known location for the stock to turn. So if I was bottom fishing for a stock to buy, I'd weigh that heavily. It's a buy signal. Realistically, however, I'd wait for price to turn and probably rise above the peak between A and B. That event would confirm an ugly double bottom. That's another buy signal, a safer, and more conservative one.
Using the numbers from the tables and chart, we can calculate the turning points.
If A (8.60) is the yearly low, then we can expect price to rise a median of 79% for a target of 15.39. Half the time price will exceed that target and half won't in about a year. The 79% value is the median A to H move for all stocks I looked at.
Using A as the reference, we should have the first leg high of 8.60 x 58% or 13.59. That corresponds to point B in the chart.
Point C, a 20% drop below B, would put the low at a minimum of 13.59 * .8 or 10.87.
Point D is a median of 18% above the peak at B or 13.59 x 18% or 16.03.
E would be 20% below D, or 12.83 (minimum).
F would be a median rise of 21% above D, or 19.40 (16.04 x 21%).
G is a minimum drop of 20% below F, or 19.41 x .8 gives 15.52.
H we already calculated as 15.39 using the median A to H rise. Using the leg approach (continuing our method), H should be 23% above F or 19.41 x 23% = 23.86.
All of these turning points are based on the yearly low, which we hope is the January 2 low. It might be better to use the November 2017 yearly low because we know it's the low for that year. It's just the wrong year.
Substituting the November low of 7.92 for 8.60 (the 2018 yearly low as of January 2) we get the following predicted turning points.
Let's see how we did by looking at how the chart unfolded in 2018.
Turn  A  B  C  D  E  F  G  H 
Predicted using 2017 low  7.92  12.51  10.01  14.77  11.81  17.87  14.29  21.98 or 14.18 
Predicted using 2 January 2018 low  8.60  13.59  10.87  16.03  12.83  19.40  15.52  23.83 or 15.39 
Actual  8.60  11.80  8.88  13.70  9.00  15.94  12.00  21.78 
The chart shows the peaks (BDFH) before 20% minimum declines (CEG). Of course, there was no guarantee that the stock would suffer three major declines along the way to H and the stock could have made a lower yearly low sometime later in the year.
The predictions for the turns are close but not dead on. However, look at turn H using the 2017 low. The predicted turn of 21.98 is just 20 cents above the actual turn. Wow! Let me hasten to add that I had no idea how these calculations would turn out. I just chose this stock and point A because it had a few 20% drops that I could use as examples.
You may wish to use the median and average numbers to form a channel, with the average numbers being the top of the channel and the median values being the bottom of the channel (or however they align).
Notice that turn F at 15.94 is priced right where the triangle's apex was (or close enough, at 16). The apex is a known support or resistance (as in this case) area. Price after F dropped 25% to bottom at G.
Support this site! Clicking any of the books (below) takes you to Amazon.com. If you buy ANYTHING while there, they pay for the referral.