The Simple Bottom Indicator signaled a ‘minor bottom’ again today (22nd September) - the first since the 16th June. It was a reasonably strong indication at 21.35% of NYSE stocks traded reaching 52w new lows (726 from 3401 as measured by the WSJ Market Data).
Be safe out there!
Platykurtic
Boilerplate recap of the ‘simple bottom detector’ rules …
[%NYSE new highs] - [%NYSE new lows] < -16.67% = Minor Bottom
[%NYSE new highs] - [%NYSE new lows] < -33.33% = Major Bottom
… where -16.67% is basically 1/6 more of the NYSE traded stocks achieved new lows than highs for a Minor Bottom & -33.33% is of course a 1/3 more (of total NYSE traded stocks) new lows than highs (as obtained from the WSJ Market Data resource). In practice this means a Major Bottom is signaled when a little more than 1/3 of the NYSE are achieving new lows.
I’m not aware of one, exactly. But something like Nasdaq recently set a new high (0 - 4ish days) and Nasdaq NH-NL < 0 (with some smoothing) can indicate a market top. Maybe works w/ other indices as well? It worked for the March 2020 (COVID) and January 2022 (current) sell-offs.
CoopersHawk, by any chance, do you have the percentages for previous bottoms (such as March 23, 2020 when S&P 500 touched a low of 2237 or March 9, 2009 when S&P 500 touched a low of 666, etc.)?
Here’s 2020 - there are no column labels but it should be self evident what the data represents (for some reason it interprets the label line from the s/sheet different to the rest of the table):
2020 table
20200226
1013
1948
79
28
352
3040
0.92105%
11.57895%
3116.39
5478110000
-10.66%
20200227
313
2700
32
22
696
3045
0.72250%
22.85714%
2978.76
7058840000
-22.13%
MinorBot
20200228
716
2308
43
8
938
3067
0.26084%
30.58363%
2954.22
8563850000
-30.32%
MinorBot
20200302
2477
542
32
16
175
3051
0.52442%
5.73582%
3090.23
6376400000
-5.21%
20200303
1031
1946
57
27
171
3034
0.88991%
5.63612%
3003.37
6355940000
-4.75%
20200304
2557
423
41
41
134
3021
1.35717%
4.43562%
3130.12
5035480000
-3.08%
20200305
327
2667
35
38
389
3029
1.25454%
12.84252%
3023.94
5575550000
-11.59%
20200306
543
2461
35
28
714
3039
0.92136%
23.49457%
2972.37
6552140000
-22.57%
MinorBot
20200309
70
2973
11
15
1620
3054
0.49116%
53.04519%
2746.56
8423050000
-52.55%
MajorBot
20200310
2304
696
47
13
688
3047
0.42665%
22.57959%
2882.23
7635960000
-22.15%
MinorBot
20200311
169
2858
22
6
1129
3049
0.19679%
37.02853%
2741.38
7374110000
-36.83%
MajorBot
20200312
67
2998
3
2
2377
3068
0.06519%
77.47718%
2480.64
8829380000
-77.41%
MajorBot
20200313
2646
376
26
3
644
3048
0.09843%
21.12861%
2711.02
8258670000
-21.03%
MinorBot
20200316
119
2934
7
0
2002
3060
0.00000%
65.42484%
2386.13
7781540000
-65.42%
MajorBot
20200317
1817
1202
31
8
1324
3050
0.26230%
43.40984%
2529.19
8358500000
-43.15%
MajorBot
20200318
186
2853
7
14
2233
3046
0.45962%
73.30926%
2398.10
8755780000
-72.85%
MajorBot
20200319
2309
724
23
5
1051
3056
0.16361%
34.39136%
2409.39
7946710000
-34.23%
MajorBot
20200320
1316
1703
36
1
299
3055
0.03273%
9.78723%
2304.92
9044690000
-9.75%
20200323
676
2360
28
2
805
3064
0.06527%
26.27285%
2237.40
7402180000
-26.21%
MinorBot
20200324
2791
244
15
1
87
3050
0.03279%
2.85246%
2447.33
7547350000
-2.82%
20200325
2422
587
26
3
54
3035
0.09885%
1.77924%
2475.56
8285670000
-1.68%
20200326
2649
369
27
4
21
3045
0.13136%
0.68966%
2630.07
7753160000
-0.56%
And here’s 2009ish:
2009ish table
20090210
463
2660
60
4
58
3183
0.12567%
1.82218%
827.16
6770170000
-1.70%
20090211
1813
1257
101
3
60
3171
0.09461%
1.89215%
833.74
5926460000
-1.80%
20090212
1422
1625
116
4
118
3163
0.12646%
3.73064%
835.19
6476460000
-3.60%
20090213
1182
1877
104
3
43
3163
0.09485%
1.35947%
826.84
5296650000
-1.26%
20090217
357
3453
64
17
331
3874
0.43882%
8.54414%
789.17
5907820000
-8.11%
20090218
897
2216
77
1
297
3190
0.03135%
9.31034%
788.42
5740710000
-9.28%
20090219
1132
2630
121
5
338
3883
0.12877%
8.70461%
778.94
5746940000
-8.58%
20090220
770
2376
56
3
555
3202
0.09369%
17.33292%
770.05
8210590000
-17.24%
MinorBot
20090223
589
3265
85
8
558
3939
0.20310%
14.16603%
743.33
6509300000
-13.96%
20090224
2703
450
51
1
322
3204
0.03121%
10.04994%
773.14
7234490000
-10.02%
20090225
1208
1908
75
6
196
3191
0.18803%
6.14228%
764.90
7483640000
-5.95%
20090226
1427
1685
78
0
171
3190
0.00000%
5.36050%
752.83
7599970000
-5.36%
20090227
1024
2065
95
1
333
3184
0.03141%
10.45854%
735.09
8926480000
-10.43%
20090302
203
2953
44
0
627
3200
0.00000%
19.59375%
700.82
7868290000
-19.59%
MinorBot
20090303
1015
2122
71
3
705
3208
0.09352%
21.97631%
696.33
7583230000
-21.88%
MinorBot
20090304
2486
644
68
1
294
3198
0.03127%
9.19325%
712.87
7673620000
-9.16%
20090305
237
2883
46
5
747
3166
0.15793%
23.59444%
682.55
7507250000
-23.44%
MinorBot
20090306
1133
1953
108
1
827
3194
0.03131%
25.89230%
683.38
7331830000
-25.86%
MinorBot
20090309
892
2231
67
2
600
3190
0.06270%
18.80878%
676.53
7277320000
-18.75%
MinorBot
20090310
2933
230
36
3
93
3199
0.09378%
2.90716%
719.60
8618330000
-2.81%
20090311
1835
1276
81
3
69
3192
0.09398%
2.16165%
721.36
7287810000
-2.07%
20090312
2864
284
39
4
61
3187
0.12551%
1.91403%
750.74
7326630000
-1.79%
20090313
1976
1118
90
4
14
3184
0.12563%
0.43970%
756.55
6787090000
-0.31%
20090316
1767
1338
73
4
16
3178
0.12587%
0.50346%
753.89
7883540000
-0.38%
20090317
2877
889
98
9
65
3864
0.23292%
1.68219%
778.12
6156800000
-1.45%
20090318
2510
601
70
6
19
3181
0.18862%
0.59730%
794.35
9098450000
-0.41%
2009 wasn’t the strongest signal strength though, most of the really strong signals occurred in 2008 during this run:
Correct in that it was almost a weak ‘Major Bottom’ by this extremely simple model. (Note I had it signal at -31.19% / -0.3119 not quite as strong as just using the lows by itself.)
Of course the various labels are just names attached to various ratio strengths, so they can be set and named however one wishes. And so I like to think of the strength of the signal as well as its name - the current signals would be very strong ‘Minor Bottom’ signals. Given this is a bear market coming from a very high overall market valuation I’d expect a few strong ‘Major Bottom’ signals before we’re finally done, but could be wrong. In that case I’d define a strong ‘Major Bottom’ as something like < -60% / -0.6 or similar. In 2008 the strongest signals were around -87% / -0.87 or so.
Thank you so much for the table. All the column headings are not really apparent, so if you could list them separately, that would be very helpful.
Also, I agree with you that the bottom detector should be viewed on a continuum, rather than in discrete/binary terms (e.g., major bottom or minor bottom). From that perspective, the current signal seems to be a very weak “major bottom” or a strong “minor bottom.”
Thank you for this information, I’m a newbie in investing right now, currently learning before starting out in real-time on a website called https://portal.tradebrains.in/ helping me to analyze stocks based on real-time and other metrics. is there any other stock screener? please do let me know about it
I have been looking for historical data for the WSJ new highs nd lows but have not been able to find historical data. Could you advise where I should look (pay?)
That’s a somewhat complicated question as the WSJ advance / decline historical data is not available on the WSJ web-site as far as I can determine. There are a couple of alternative options depending on what’s required …
If one just wants to obtain the data required for the Simple Bottom Detector then just drop me a note and I can supply it in s/sheet form back to 2005ish.
More comprehensive data is available from 2005ish to 10 February 2020 at Advance Decline data. This is where I obtained the original (NYSE advance / decline) data-set and have kept it up-to-date since then.
The recommendation when unicorn stopped updating its data set was to transition to Pinnacle Data at https://pinnacledata2.com/. It costs - I’m not sure about current pricing but it wasn’t too expensive last I looked and you’d need to validate it against the WSJ market data - but (from memory) it also extends back further and frankly the Simple Bottom Detector is so simple that it’s probably plug-and-play as the slight-to-moderate differences between data sets isn’t likely at all material. So Pinnacle was where I was planning to look if I needed a greater or longer data set.
Thanks for the detailed response, very helpful. I looked at the Pinnacle Data and I think I will go down that path. The data goes back to 1965 for NYSE.
@aussi - If you look at the old boards, I think you can get my name. You can then search and find an email for me (I’m a mathematician in Georgia). Feel free to email me. I’ve written code that may help you process the data. I’d be happy to share.
Try Barchart.com. A free account brings a lot of downloadable data including daily “NYSE” advances, declines and unchanged statistics. They have their own composite indices such as $ADVT, $DECT, $UNCT as well.
The Barchart data does not match the WSJ data in total number of companies, new highs and new lows. My guess is that this is probably not an issue, I am just trying to tag along with platykurtic’s work.