1 00:00:00,000 --> 00:00:10,920 Welcome to the Cannabis Data Science Meetup Group. 2 00:00:10,920 --> 00:00:12,400 Should be a really good day today. 3 00:00:12,400 --> 00:00:16,520 We're going to pick up with looking at market performance. 4 00:00:16,520 --> 00:00:21,280 We've got a good crowd today, so welcome to the group. 5 00:00:21,280 --> 00:00:29,560 So normally I go on a spiel about economics, data science, and the cannabis industry. 6 00:00:29,560 --> 00:00:31,880 We've got some new people here today. 7 00:00:31,880 --> 00:00:38,600 So if you're interested, we could do a round of introductions either now or at the end. 8 00:00:38,600 --> 00:00:45,840 So Andrew and our new guest, you're welcome to chime in and introduce yourselves because 9 00:00:45,840 --> 00:00:54,160 we'd love to do a nice back and forth if you have any questions or want to steer the conversation 10 00:00:54,160 --> 00:00:55,160 in any way. 11 00:00:55,160 --> 00:00:56,160 Sure. 12 00:00:56,160 --> 00:00:58,800 Hi everyone, I'm Andrew. 13 00:00:58,800 --> 00:01:03,400 I'm just a student working on a master's in data science. 14 00:01:03,400 --> 00:01:07,680 I was just looking forward to things to do and I saw this meetup and it fit my schedule, 15 00:01:07,680 --> 00:01:10,680 so I decided to hop on in. 16 00:01:10,680 --> 00:01:14,240 Well, awesome to have you, Andrew. 17 00:01:14,240 --> 00:01:18,480 Always have a lot of people interested in data science. 18 00:01:18,480 --> 00:01:25,960 So feel free to chime in at any point if you have any questions or ideas or anything you 19 00:01:25,960 --> 00:01:33,120 want to talk about. 20 00:01:33,120 --> 00:01:50,560 Without further ado, I'll go ahead and share with you what I've prepared today. 21 00:01:50,560 --> 00:01:57,120 So we've been looking at predicting market performance. 22 00:01:57,120 --> 00:02:06,960 So this is the data and forecasts we were looking at and preparing last week. 23 00:02:06,960 --> 00:02:16,720 So we were looking at cannabis sales here in Massachusetts, which we saw are going up 24 00:02:16,720 --> 00:02:17,720 and up and up. 25 00:02:17,720 --> 00:02:22,480 However, there have been turbulent periods. 26 00:02:22,480 --> 00:02:33,240 So we'll go into it, but our data could be biased by these past turbulent events. 27 00:02:33,240 --> 00:02:44,760 So for example, this period, this dip in 2021 in January of 2021 was used in our forecasts. 28 00:02:44,760 --> 00:02:50,160 And as you can see, our forecasts for 2022 also dip in January. 29 00:02:50,160 --> 00:02:57,920 So if you're interested in doing forecasts of your own, it could be interesting to limit 30 00:02:57,920 --> 00:03:04,280 your forecast training data to perhaps just the past six months. 31 00:03:04,280 --> 00:03:09,760 You're going to have really wide confidence bounds because you're going to have only a 32 00:03:09,760 --> 00:03:15,000 handful of observations, well, maybe more than a handful. 33 00:03:15,000 --> 00:03:17,800 You could have a couple dozen if you're doing weekly. 34 00:03:17,800 --> 00:03:24,920 So you could have a number of observations there that could be more representative. 35 00:03:24,920 --> 00:03:32,680 And so this is where we were starting to realize that there's a lot of assumptions being made 36 00:03:32,680 --> 00:03:35,280 when you start to forecast. 37 00:03:35,280 --> 00:03:45,640 So we want to dive into some of these assumptions today and want to spend a bit of time today 38 00:03:45,640 --> 00:03:52,480 on some of the economic theory to explain why we're looking at these data points. 39 00:03:52,480 --> 00:04:04,320 So it makes sense why we may want to look at sales and plants grown here in Massachusetts. 40 00:04:04,320 --> 00:04:10,160 We want to look at the retail side and the wholesale side, see how the producers are 41 00:04:10,160 --> 00:04:13,320 doing. 42 00:04:13,320 --> 00:04:22,520 And so we looked at, OK, how many retailers are there in Massachusetts and how many cultivators 43 00:04:22,520 --> 00:04:25,840 are there in Massachusetts? 44 00:04:25,840 --> 00:04:40,960 OK, bear with me. 45 00:04:40,960 --> 00:04:57,520 We're going to use a different presentation and style next week, so we gave this one its 46 00:04:57,520 --> 00:04:59,160 chance. 47 00:04:59,160 --> 00:05:07,360 OK, so we're looking at retailers and cultivators. 48 00:05:07,360 --> 00:05:11,560 And we did a forecast for each. 49 00:05:11,560 --> 00:05:24,040 Note, these are difficult series to predict because there's actually not a lot of variability 50 00:05:24,040 --> 00:05:26,080 in the series themselves. 51 00:05:26,080 --> 00:05:32,180 So there could be other underlying factors which one would expect explaining the number 52 00:05:32,180 --> 00:05:34,840 of retailers and the number of cultivators. 53 00:05:34,840 --> 00:05:46,040 So this depends largely on the Cannabis Commission in Massachusetts, their licensing process, 54 00:05:46,040 --> 00:05:47,400 who they choose to license. 55 00:05:47,400 --> 00:05:53,560 Of course, it does matter to a certain extent on how many applicants are being submitted, 56 00:05:53,560 --> 00:05:57,600 how many applications are being submitted. 57 00:05:57,600 --> 00:06:05,400 Our prediction, crude as it may be, predicts that there may be a modest increase in retailers 58 00:06:05,400 --> 00:06:09,080 or potentially even a modest decrease. 59 00:06:09,080 --> 00:06:19,120 So we're predicting anywhere between around 350 to around 420, 2025 retailers or so in 60 00:06:19,120 --> 00:06:23,960 the market by the end of 2022. 61 00:06:23,960 --> 00:06:30,440 And right now there's around 375, 380 retailers. 62 00:06:30,440 --> 00:06:33,440 And we... 63 00:06:33,440 --> 00:06:37,320 Oh no. 64 00:06:37,320 --> 00:06:40,800 OK. 65 00:06:40,800 --> 00:06:45,800 So we're having... 66 00:06:45,800 --> 00:06:54,560 It does not like the zoom. 67 00:06:54,560 --> 00:07:07,400 So with cultivators, we're actually predicting a slightly higher growth in retailers. 68 00:07:07,400 --> 00:07:14,960 So I forget the precise number, but we're hovering between 250 and 300 at the moment, 69 00:07:14,960 --> 00:07:18,680 which is close to 300. 70 00:07:18,680 --> 00:07:25,520 And we're predicting, as compared to retailers, higher growth and cultivation. 71 00:07:25,520 --> 00:07:33,880 So we'll predict that at the higher end, you may see around 350 cultivators. 72 00:07:33,880 --> 00:07:41,040 In Massachusetts, by the end of 2022, and on the lower end, we're still predicting around 73 00:07:41,040 --> 00:07:42,040 300. 74 00:07:42,040 --> 00:07:48,920 So we're predicting a growth in cultivation one way or the other. 75 00:07:48,920 --> 00:07:50,800 So we'll see if that pans out. 76 00:07:50,800 --> 00:07:55,960 And like we said, a lot of this depends on the regulation. 77 00:07:55,960 --> 00:08:01,200 We've also got... 78 00:08:01,200 --> 00:08:04,200 We've also got... 79 00:08:04,200 --> 00:08:06,200 OK. 80 00:08:06,200 --> 00:08:17,240 Pardon me for this poor presentation here. 81 00:08:17,240 --> 00:08:23,840 Like we said, we're going to use a different presenting software next week. 82 00:08:23,840 --> 00:08:27,120 All right. 83 00:08:27,120 --> 00:08:29,120 So we're going to... 84 00:08:29,120 --> 00:08:34,800 We're looking at sales, number of retailers, looking at plants and cultivators. 85 00:08:34,800 --> 00:08:43,520 Well, we realized we can create a metric that perhaps not many people have looked at. 86 00:08:43,520 --> 00:08:48,120 And it can provide some nice insights for us. 87 00:08:48,120 --> 00:08:53,280 So we've got cannabis sales, we've got retailers. 88 00:08:53,280 --> 00:08:56,000 Well, let's create a new statistic. 89 00:08:56,000 --> 00:08:59,160 Average sales per retailer. 90 00:08:59,160 --> 00:09:09,400 So it would be awesome if we actually had sales data for each of the retailers. 91 00:09:09,400 --> 00:09:11,960 And the data exists out there. 92 00:09:11,960 --> 00:09:15,000 I'm sure the state of Massachusetts has that data. 93 00:09:15,000 --> 00:09:18,680 However, we'll make do with what we have. 94 00:09:18,680 --> 00:09:27,720 And so we can just divide the average sales by retailers here. 95 00:09:27,720 --> 00:09:44,680 Control L. 96 00:09:44,680 --> 00:09:48,000 And this is going to be an interesting metric. 97 00:09:48,000 --> 00:09:52,920 And I'll explain a bit more further on. 98 00:09:52,920 --> 00:10:02,680 This is going to be our first metric where we sort of start to get at market performance. 99 00:10:02,680 --> 00:10:10,160 So how is the market performing for these parties that are in the market? 100 00:10:10,160 --> 00:10:11,800 And what are some of the parties? 101 00:10:11,800 --> 00:10:16,920 Well, there's the retailers and the cultivators, or two of them. 102 00:10:16,920 --> 00:10:17,920 Right. 103 00:10:17,920 --> 00:10:22,200 And so how are these retailers doing over time? 104 00:10:22,200 --> 00:10:36,800 And so we see that in the beginning, you have modest weekly sales at retailers. 105 00:10:36,800 --> 00:10:43,080 So it starts around 20,000 per week per retailer on average. 106 00:10:43,080 --> 00:10:46,840 Some are obviously above average, some are below average. 107 00:10:46,840 --> 00:10:47,840 And this is climbing. 108 00:10:47,840 --> 00:10:48,840 Right. 109 00:10:48,840 --> 00:10:55,040 And you get around 50,000 per week per retailer. 110 00:10:55,040 --> 00:11:06,400 Then of course, retail closes in Massachusetts for a two month span in April and May of 2020. 111 00:11:06,400 --> 00:11:13,240 Average sales per retailer takes a hit when they reopen. 112 00:11:13,240 --> 00:11:19,360 So when they reopen, they're below the 50K a week. 113 00:11:19,360 --> 00:11:26,160 However, they get back up to the 50K in just a matter of a couple of weeks. 114 00:11:26,160 --> 00:11:32,600 And then once again, you're on a modest trajectory up. 115 00:11:32,600 --> 00:11:43,440 Once again, rocky boat in spring, well, I guess this would be winter slash early spring 116 00:11:43,440 --> 00:11:45,640 of 2021. 117 00:11:45,640 --> 00:11:50,560 You have a rocky time in the market here in Massachusetts. 118 00:11:50,560 --> 00:11:52,880 Like I'm going to have to do a bit more historical digging. 119 00:11:52,880 --> 00:11:53,880 Right. 120 00:11:53,880 --> 00:11:55,920 And so this is what's so cool about economics. 121 00:11:55,920 --> 00:11:56,920 Right. 122 00:11:56,920 --> 00:12:05,040 Because not only are you studying economic theory, you're studying statistics, but you're 123 00:12:05,040 --> 00:12:08,160 also a student of history. 124 00:12:08,160 --> 00:12:09,160 Right. 125 00:12:09,160 --> 00:12:18,280 Because we need to look in the history books, in the newspapers and find out what was happening 126 00:12:18,280 --> 00:12:19,560 at this time. 127 00:12:19,560 --> 00:12:23,440 In specific, what was happening in Massachusetts at this time. 128 00:12:23,440 --> 00:12:39,280 So if you were, I want to use a better word than entrepreneurial, but if you're a go getter, 129 00:12:39,280 --> 00:12:48,120 then you can go hit the local Massachusetts newspapers in this time period and try to 130 00:12:48,120 --> 00:12:52,720 figure out, just kind of like look at the local stories and kind of see what's happening. 131 00:12:52,720 --> 00:12:54,560 So this is where, right. 132 00:12:54,560 --> 00:12:57,600 So they're not going to talk on a national scale. 133 00:12:57,600 --> 00:13:03,440 They're not probably not going to talk in the national newspapers about cannabis markets 134 00:13:03,440 --> 00:13:06,240 being closed in Massachusetts. 135 00:13:06,240 --> 00:13:07,240 Right. 136 00:13:07,240 --> 00:13:15,080 But right, if you look in the local newspapers or this or that, there could have been townships 137 00:13:15,080 --> 00:13:24,840 or counties where businesses may have locked down or I'm just conjecturing, who knows what 138 00:13:24,840 --> 00:13:25,840 happened. 139 00:13:25,840 --> 00:13:34,000 There could have been the hots, late in Viroid may have hit particularly hard or who knows. 140 00:13:34,000 --> 00:13:38,560 There could have been a tourism drought or there's a lot of factors. 141 00:13:38,560 --> 00:13:43,320 So long story short, check out the history books and then try to incorporate history 142 00:13:43,320 --> 00:13:49,620 into your analysis. 143 00:13:49,620 --> 00:13:56,440 Our main takeaway was average sales per retailer. 144 00:13:56,440 --> 00:14:02,600 It's actually, it's continued to increase and it's getting substantial, right. 145 00:14:02,600 --> 00:14:10,880 You've got average sales around 75 or so thousand a week. 146 00:14:10,880 --> 00:14:12,680 You know, not shabby. 147 00:14:12,680 --> 00:14:18,620 And this is where if you're a student of economics, it would be worthwhile to compare this to 148 00:14:18,620 --> 00:14:20,160 other industries. 149 00:14:20,160 --> 00:14:26,160 So what's the average sale in the restaurant industry? 150 00:14:26,160 --> 00:14:34,840 What's the average sales for, you know, an agricultural producer of another agricultural 151 00:14:34,840 --> 00:14:39,480 crop? 152 00:14:39,480 --> 00:14:45,360 So you can start comparing cannabis to other industries here. 153 00:14:45,360 --> 00:14:49,800 But what we noticed and what we would like to measure, and if you tune into Saturday 154 00:14:49,800 --> 00:14:58,240 morning statistics, we do this more rigorously, but you can actually see if there's a change 155 00:14:58,240 --> 00:15:04,840 in factors or perhaps volatility during this period, right. 156 00:15:04,840 --> 00:15:08,800 So there may have been an underlying structural change. 157 00:15:08,800 --> 00:15:17,800 And what I notice is this period is much more volatile than the prior periods, right. 158 00:15:17,800 --> 00:15:28,160 So especially 2019 to 2020, this, there was just a fairly steady period of growth, right. 159 00:15:28,160 --> 00:15:33,240 I'm sure during that time they felt like things were tumultuous, but they're sure, they're 160 00:15:33,240 --> 00:15:40,760 sure not as tumultuous, arguably as 2021. 161 00:15:40,760 --> 00:15:44,160 So like, look at this volatility you have here. 162 00:15:44,160 --> 00:15:51,720 It's like, yeah, average sales may be high or on average, but you have incredible volatility. 163 00:15:51,720 --> 00:15:58,520 This may make it difficult for your general managers, may make it tricky to stock inventory 164 00:15:58,520 --> 00:15:59,640 correctly, right. 165 00:15:59,640 --> 00:16:07,280 If you're having sales peak one week and then they drop off the cliff the next week, well, 166 00:16:07,280 --> 00:16:13,680 you know, you may overstock your stores, you may have your stores understocked when sales 167 00:16:13,680 --> 00:16:18,040 spike. 168 00:16:18,040 --> 00:16:22,760 So that's why it leads to the importance of forecasting. 169 00:16:22,760 --> 00:16:25,880 And here we did a medium term forecast. 170 00:16:25,880 --> 00:16:31,120 And like we, I think I brought this up in Saturday morning statistics, but if you're 171 00:16:31,120 --> 00:16:38,320 a general manager, it would be helpful to do a in depth, say 30 day forecast. 172 00:16:38,320 --> 00:16:40,640 So here we used weekly data. 173 00:16:40,640 --> 00:16:48,320 Your general manager, it could be useful to just use daily data and just do a 30 day forecast. 174 00:16:48,320 --> 00:16:57,160 So that way you can use day of the week effects and get a real nice prediction for the coming 175 00:16:57,160 --> 00:16:58,160 month. 176 00:16:58,160 --> 00:17:04,480 But here we just looked at the coming year because we're interested in kind of the bigger 177 00:17:04,480 --> 00:17:10,720 scheme trajectory here in the Canvas data science meetup group, right. 178 00:17:10,720 --> 00:17:15,960 So we've been looking at the, we're trying to look at the performance of these industries. 179 00:17:15,960 --> 00:17:22,280 And so what we're noticing is the retailers are performing, one would argue better and 180 00:17:22,280 --> 00:17:28,840 better, but kind of we've got some volatility issues that may need to get ironed out. 181 00:17:28,840 --> 00:17:44,360 Okay, let's see if we can. 182 00:17:44,360 --> 00:17:55,160 For the cultivation, cultivation is actually kind of got a different story going on where 183 00:17:55,160 --> 00:18:00,400 you've got cultivation getting up to speed slowly. 184 00:18:00,400 --> 00:18:04,040 And this is actually kind of what you see in a lot of cannabis markets, right. 185 00:18:04,040 --> 00:18:10,360 So it's going to take your cultivation, you know, a good six months or so to get up to 186 00:18:10,360 --> 00:18:12,800 speed for the market. 187 00:18:12,800 --> 00:18:17,600 So this is essentially people getting plants in the ground. 188 00:18:17,600 --> 00:18:25,720 So you don't, you know, it takes time to get these facilities built, get your doors opened, 189 00:18:25,720 --> 00:18:33,480 get the lights on, get the employees hired, you know, get your HVAC system installed. 190 00:18:33,480 --> 00:18:37,200 You know, there's a lot of preparation, get your permits, right. 191 00:18:37,200 --> 00:18:39,640 So it's a whole long game, right. 192 00:18:39,640 --> 00:18:46,320 So people think, oh, I'm just going to get my license and, you know, start growing a 193 00:18:46,320 --> 00:18:48,120 thousand plants tomorrow. 194 00:18:48,120 --> 00:18:51,000 Well, that's not, that's not the case. 195 00:18:51,000 --> 00:18:54,460 You've got, you've got to jump through a lot of hoops. 196 00:18:54,460 --> 00:18:58,200 You've got to get your local permits, you know, you've got to make sure your license 197 00:18:58,200 --> 00:18:59,280 is locked in. 198 00:18:59,280 --> 00:19:04,600 You have to get your employees, you have to get all your capital equipment, capital equipment 199 00:19:04,600 --> 00:19:07,880 may require funding, right. 200 00:19:07,880 --> 00:19:12,520 And the funding isn't entirely different. 201 00:19:12,520 --> 00:19:20,200 Candle worms in the cannabis industry as it is in other industries, because, you know, 202 00:19:20,200 --> 00:19:27,360 you can't just get conventional loans from large banks like you may be able to in other 203 00:19:27,360 --> 00:19:30,600 agricultural industries. 204 00:19:30,600 --> 00:19:37,640 Long story short, takes cultivation a little while to get what I would call up to speed, 205 00:19:37,640 --> 00:19:38,640 right. 206 00:19:38,640 --> 00:19:44,520 So here you just see this exponential growth in the number of plants in the ground. 207 00:19:44,520 --> 00:19:46,500 And this is the tract number of plants. 208 00:19:46,500 --> 00:19:50,520 You can break this down by vegetative and flowering. 209 00:19:50,520 --> 00:19:55,840 So if you've got more insight than I do, then by all means, you can look at the data slightly 210 00:19:55,840 --> 00:19:56,840 differently. 211 00:19:56,840 --> 00:20:05,680 But in general, the trajectory of the number of plants is the same or similar. 212 00:20:05,680 --> 00:20:14,680 Long story short, plants get up to around this, you know, steady level of 500 to around 213 00:20:14,680 --> 00:20:18,800 600 plants per cultivator. 214 00:20:18,800 --> 00:20:20,500 And that's quite steady. 215 00:20:20,500 --> 00:20:29,680 So you do see what you could probably estimate, you know, a modest trend in that data. 216 00:20:29,680 --> 00:20:38,800 But once again, you've got, you know, maybe minor cyclical business cycle trends going 217 00:20:38,800 --> 00:20:40,480 on. 218 00:20:40,480 --> 00:20:52,180 But even with what's interesting, right, is there's not a dramatic effect on the number 219 00:20:52,180 --> 00:21:00,380 of plants in 2020 or really in January of 2021. 220 00:21:00,380 --> 00:21:07,320 So the cultivators, and this would make sense, right, if you've got your cultivation set 221 00:21:07,320 --> 00:21:13,280 up, you've got your plants in the ground, you know, it's not even really going to be 222 00:21:13,280 --> 00:21:17,800 that big of a factor if, if retail is closed. 223 00:21:17,800 --> 00:21:21,320 And in fact, you even see a little bit of an uptick. 224 00:21:21,320 --> 00:21:28,440 Maybe people even took that took that time to get even more plants in the ground. 225 00:21:28,440 --> 00:21:35,640 So you know, the plants per cultivator, so the cultivators, their production, it looks 226 00:21:35,640 --> 00:21:37,960 like stayed steady. 227 00:21:37,960 --> 00:21:46,520 And in fact, it looks like they realized they needed to turn on the gas in spring of 2021. 228 00:21:46,520 --> 00:21:55,120 And then this makes me think that there's all of a sudden an increase in demand in the 229 00:21:55,120 --> 00:21:57,260 Massachusetts market. 230 00:21:57,260 --> 00:22:04,760 And so once again, time to hit the history books, time to hit the newspapers, what was 231 00:22:04,760 --> 00:22:16,200 going on in Massachusetts in mid to late 2021. 232 00:22:16,200 --> 00:22:20,760 Really interesting to see and you know, see if there's a lot of articles talking about 233 00:22:20,760 --> 00:22:23,880 cannabis cultivation. 234 00:22:23,880 --> 00:22:33,160 So once again, you see this large growth in the number of plants per cultivator, average 235 00:22:33,160 --> 00:22:34,160 number. 236 00:22:34,160 --> 00:22:39,960 And so once again, hedge this in that, right, there could be some cultivators that just have 237 00:22:39,960 --> 00:22:45,460 a ginormous amount of plants and some that don't have very many. 238 00:22:45,460 --> 00:22:53,120 So this could just be the entrance of mega growers, you know, that's not impossible. 239 00:22:53,120 --> 00:23:00,440 That just like some mega growers just came online here in late 2021. 240 00:23:00,440 --> 00:23:02,720 That's not impossible. 241 00:23:02,720 --> 00:23:11,480 However, the story I generally been telling is that it looks like cultivators, that sort 242 00:23:11,480 --> 00:23:18,440 of been exploiting economies of scale, there's increased demand, they're growing more plants. 243 00:23:18,440 --> 00:23:25,720 However, it looks like they may have kind of reached a peak and maybe cutting back a 244 00:23:25,720 --> 00:23:27,480 little bit. 245 00:23:27,480 --> 00:23:37,920 So I think this is going to be one of the hardest series to predict coincidentally because 246 00:23:37,920 --> 00:23:42,760 we don't it looks like things are changing kind of recently, right? 247 00:23:42,760 --> 00:23:48,520 It looks like really, in the past month or two, there's been almost like a structural 248 00:23:48,520 --> 00:23:53,200 change here in cultivation. 249 00:23:53,200 --> 00:23:58,120 So the other series will same for plants. 250 00:23:58,120 --> 00:24:04,740 So you know, the other series had their own structural breaks. 251 00:24:04,740 --> 00:24:16,400 But for whatever reason, there's something going on in the cultivation side of the market. 252 00:24:16,400 --> 00:24:22,380 And that's going to be interesting to see what happens in 2022. 253 00:24:22,380 --> 00:24:31,000 Our forecasts because there's such a dip, we predict there's going to be a large dip 254 00:24:31,000 --> 00:24:32,920 in 2022. 255 00:24:32,920 --> 00:24:38,880 I would love to rework these forecasts and not have such a dramatic dip because just 256 00:24:38,880 --> 00:24:46,880 my Bayesian prior is that there's not going to be this dramatic dip in January, like I 257 00:24:46,880 --> 00:24:52,920 predict a modest dip, but not one that dramatic. 258 00:24:52,920 --> 00:25:00,720 And this is what's cool about Bayesian statistics is you acknowledge your prior biases. 259 00:25:00,720 --> 00:25:07,960 And you can kind of use them to make better forecasts if you want to boil it down that 260 00:25:07,960 --> 00:25:08,960 way. 261 00:25:08,960 --> 00:25:17,480 But tune into Saturday morning statistics and you can get a lot about the nerdy aspects 262 00:25:17,480 --> 00:25:21,160 of the statistics. 263 00:25:21,160 --> 00:25:27,800 Long story short, our predictions are that the average plants per cultivator is going 264 00:25:27,800 --> 00:25:36,960 to fall back to its pre 2021 level, where you're going to see around 600 plants per 265 00:25:36,960 --> 00:25:38,680 cultivator. 266 00:25:38,680 --> 00:25:41,560 I think this is going to be one of like I said, I think this is going to be one of the 267 00:25:41,560 --> 00:25:54,160 more interesting series to follow because right, the supply of cannabis is going to 268 00:25:54,160 --> 00:25:58,320 be a major determination of the performance of the market rate. 269 00:25:58,320 --> 00:26:04,720 So that's going to determine a lot of the set of the prices. 270 00:26:04,720 --> 00:26:13,200 And, and essentially the profits for the retailers and the cultivators in a surplus that will 271 00:26:13,200 --> 00:26:17,160 be had by the consumers. 272 00:26:17,160 --> 00:26:25,240 Talking about profits and consumers, well, let's go ahead and do a little bit of a history 273 00:26:25,240 --> 00:26:34,080 slash economics lesson, because we've been neglecting the economics that's underlying 274 00:26:34,080 --> 00:26:36,120 our analysis. 275 00:26:36,120 --> 00:26:41,780 And I wanted to give you a bit of a history lesson about where this all began. 276 00:26:41,780 --> 00:26:48,840 Some of the people that pioneered the work pioneered sort of the amount type of analysis 277 00:26:48,840 --> 00:26:54,960 that we're going to be doing and what people look at. 278 00:26:54,960 --> 00:27:00,600 We don't have to look at these things necessarily, but it's just interesting to look at the history 279 00:27:00,600 --> 00:27:06,240 of of where these tools came about from. 280 00:27:06,240 --> 00:27:12,760 So long story short, you have a professor, Edward S. Mason at Harvard University. 281 00:27:12,760 --> 00:27:19,440 He's the first one who's really extending upon sort of the theory of the firm. 282 00:27:19,440 --> 00:27:22,240 And what's the theory of the firm? 283 00:27:22,240 --> 00:27:31,020 Well, this is where you sort of determine, okay, you know, what gets produced internally 284 00:27:31,020 --> 00:27:36,400 at an organization and what gets bought and sold on the open market. 285 00:27:36,400 --> 00:27:37,400 Right. 286 00:27:37,400 --> 00:27:44,720 So sometimes firms outsource and sometimes they do things in house. 287 00:27:44,720 --> 00:27:46,400 Right. 288 00:27:46,400 --> 00:27:53,240 And a lot of times economics sort of ignores, you know, things that are going on in house. 289 00:27:53,240 --> 00:27:54,240 Right. 290 00:27:54,240 --> 00:27:58,860 Because we're just looking at a lot of times economics is just looking at transactions 291 00:27:58,860 --> 00:27:59,860 between parties. 292 00:27:59,860 --> 00:28:06,320 So you're just looking at, say, producer A buying and selling it from producer B. 293 00:28:06,320 --> 00:28:14,600 Well, you don't really take into consideration that, you know, producer A may be taking actions 294 00:28:14,600 --> 00:28:20,720 so that they can then internally produce what producer B is producing and just do things 295 00:28:20,720 --> 00:28:21,720 in house. 296 00:28:21,720 --> 00:28:23,920 And you see companies do this. 297 00:28:23,920 --> 00:28:33,400 So you see acquisitions, you see this a lot with the large technology companies where 298 00:28:33,400 --> 00:28:44,680 large technology company A will acquire small, you know, small technology company B and then 299 00:28:44,680 --> 00:28:47,680 just sort of bring that in house. 300 00:28:47,680 --> 00:28:53,260 So that's the theory of the firm has a lot to do about costs. 301 00:28:53,260 --> 00:29:00,320 So what's the cost of doing things internally versus buying it on the market? 302 00:29:00,320 --> 00:29:08,080 Well, this leads to, okay, well, what's the layout of the market? 303 00:29:08,080 --> 00:29:16,120 You know, what are the size of these players in the market and how do they conduct themselves 304 00:29:16,120 --> 00:29:21,180 and what's the outcome, aka the performance? 305 00:29:21,180 --> 00:29:31,160 So we're looking at how things are structured, how firms behave and people, and then what's 306 00:29:31,160 --> 00:29:32,740 the outcome? 307 00:29:32,740 --> 00:29:38,880 How do the firms perform and what's the consequence on like society as a whole? 308 00:29:38,880 --> 00:29:39,880 So consumers. 309 00:29:39,880 --> 00:29:47,520 And then that's where we sort of get, I'll talk a bit more about Joseph Bain and George 310 00:29:47,520 --> 00:29:58,040 Stigler coming up, but those are two economists who really led the work forward in a similar 311 00:29:58,040 --> 00:30:00,840 vein as we've been doing. 312 00:30:00,840 --> 00:30:02,840 Where this is led? 313 00:30:02,840 --> 00:30:10,560 Well, it's essentially led to antitrust policies, which, you know, may not have come about otherwise. 314 00:30:10,560 --> 00:30:26,760 And so what they noticed was, okay, as they argued, markets become more concentrated. 315 00:30:26,760 --> 00:30:36,700 Well, if there's only a small number of players, it's going to essentially be cheaper to collude. 316 00:30:36,700 --> 00:30:42,640 So let's say there's a fixed cost, so you have to pay people off to collude. 317 00:30:42,640 --> 00:30:48,460 Well, the more people there are, the more people you have to pay off, thus the higher 318 00:30:48,460 --> 00:30:50,980 the cost of collusion. 319 00:30:50,980 --> 00:31:00,440 So the argument is, the more concentrated the market is, the more likely collusion is 320 00:31:00,440 --> 00:31:04,640 to occur because there's a lower cost. 321 00:31:04,640 --> 00:31:09,780 And if collusion occurs, that can have an adverse effect on society. 322 00:31:09,780 --> 00:31:16,840 So essentially, consumers are going to face a higher price and lose consumer surplus. 323 00:31:16,840 --> 00:31:20,840 Of course, there's a counter argument, right? 324 00:31:20,840 --> 00:31:25,720 The counter argument is, well, if you're naturally efficient, right? 325 00:31:25,720 --> 00:31:30,820 So if you do things better than other people in the industry, well, you're just going to 326 00:31:30,820 --> 00:31:37,980 naturally gain a market share, and just because you have a large market share, that doesn't 327 00:31:37,980 --> 00:31:41,440 mean you're going to collude, right? 328 00:31:41,440 --> 00:31:50,580 Just because you're the only or one of the only players in a market, that doesn't dictate 329 00:31:50,580 --> 00:31:54,500 that you have to collude with people. 330 00:31:54,500 --> 00:32:01,120 So that's sort of the counter argument, the argument is, well, that's inevitably what 331 00:32:01,120 --> 00:32:02,120 we see. 332 00:32:02,120 --> 00:32:06,820 And this is sort of where you kind of can get into psychology of groups. 333 00:32:06,820 --> 00:32:12,900 And once again, economics, we like the dabble, and of course, you have to look at the psychology 334 00:32:12,900 --> 00:32:14,900 of these agents. 335 00:32:14,900 --> 00:32:17,300 And I think it's real interesting, right? 336 00:32:17,300 --> 00:32:24,660 And so that's where psychologists can kind of help out is, okay, like, you know, do you 337 00:32:24,660 --> 00:32:34,100 see like bad behavior slash collusion in, you know, small groups versus large groups? 338 00:32:34,100 --> 00:32:40,980 And just my personal belief is, you know, the more people you add to a group, the harder 339 00:32:40,980 --> 00:32:47,120 it is to say collude and do bad actions, right? 340 00:32:47,120 --> 00:32:53,420 Because there's more people, you don't know which ones are going to turn you in. 341 00:32:53,420 --> 00:33:00,340 I don't know. 342 00:33:00,340 --> 00:33:03,540 And so I think there's more work that can be studied there. 343 00:33:03,540 --> 00:33:13,420 But that's sort of, that's a major pinning point underlying market concentration analysis 344 00:33:13,420 --> 00:33:20,300 is because the implicit argument you're making when you're measuring market concentration 345 00:33:20,300 --> 00:33:27,620 is that an increase in market concentration could be bad for consumers. 346 00:33:27,620 --> 00:33:36,580 Meanwhile, you know, a lower degree of market concentration could be beneficial for consumers. 347 00:33:36,580 --> 00:33:44,380 Like I said, the counter argument is if you actively work to curtail concentration, then 348 00:33:44,380 --> 00:33:52,860 you could be curtailing efficiencies that would otherwise been gained. 349 00:33:52,860 --> 00:33:58,580 So the idea is you don't necessarily want to limit concentration just for the sake of 350 00:33:58,580 --> 00:34:09,740 limiting it, but you may want to keep an eye out on concentrated markets because it may 351 00:34:09,740 --> 00:34:18,860 be ripe for collusion. 352 00:34:18,860 --> 00:34:26,460 And this was a lot of the work that was done by Bain, Joseph Bain. 353 00:34:26,460 --> 00:34:34,220 Stigler looked a lot at regulatory capture. 354 00:34:34,220 --> 00:34:45,820 And this is, of course, Stigler taught partially at the University of Chicago. 355 00:34:45,820 --> 00:34:53,740 So this is quite a University of Chicago type of stance on regulation. 356 00:34:53,740 --> 00:34:57,380 So just kind of keep that in mind. 357 00:34:57,380 --> 00:35:06,100 University of Chicago school economists are typically not in favor of overarching regulations. 358 00:35:06,100 --> 00:35:11,580 That's sort of a generalization, and it may be more true historically than present day, 359 00:35:11,580 --> 00:35:15,580 but that's what you see a lot of with these economists. 360 00:35:15,580 --> 00:35:20,900 They make good arguments, and so you can't just ignore them. 361 00:35:20,900 --> 00:35:24,220 So that's sort of the point about history. 362 00:35:24,220 --> 00:35:32,060 You have to study the people and the points they make, whether you agree with them or 363 00:35:32,060 --> 00:35:33,060 not. 364 00:35:33,060 --> 00:35:37,560 These were people that led a lot of the groundwork. 365 00:35:37,560 --> 00:35:44,300 So it's just interesting and arguably important to study them. 366 00:35:44,300 --> 00:35:49,820 That way you kind of know what you're talking about when we look. 367 00:35:49,820 --> 00:35:58,160 So that way when we're talking about market performance in the cannabis industry, we actually 368 00:35:58,160 --> 00:36:04,900 know about the history of the first people who studied market performance. 369 00:36:04,900 --> 00:36:11,260 And one of these people was George Stigler. 370 00:36:11,260 --> 00:36:20,100 And his argument was you have to watch out for essentially regulation for regulation's 371 00:36:20,100 --> 00:36:30,060 sake, which kind of is part of this counter argument here where you don't want to just 372 00:36:30,060 --> 00:36:39,460 curtail concentration for concentration's sake. 373 00:36:39,460 --> 00:36:49,660 And what Stigler argued is the regulatory agencies, so in this case it would be the 374 00:36:49,660 --> 00:36:54,900 Massachusetts Cannabis Commission, they can be captured. 375 00:36:54,900 --> 00:37:05,100 So this would be lobbyists kind of persuading the government bodies to essentially enact 376 00:37:05,100 --> 00:37:08,260 regulations at their behest. 377 00:37:08,260 --> 00:37:13,980 So they're sort of nudging the politicians so they may take them out to lunch and say, 378 00:37:13,980 --> 00:37:20,420 oh, don't you think it would be a good idea if we did X, Y, and Z? 379 00:37:20,420 --> 00:37:29,380 Well X, Y, and Z may be easy for that producer to do, but it may be tough for their competition 380 00:37:29,380 --> 00:37:30,380 to do. 381 00:37:30,380 --> 00:37:43,700 So next thing you know, we've got a regulation that, so for example, there's a regulation 382 00:37:43,700 --> 00:37:49,460 for delivery that you have to have four cameras on your delivery vehicle or something like 383 00:37:49,460 --> 00:37:55,620 that. 384 00:37:55,620 --> 00:38:02,380 But it does seem like one could argue, yes, that is for the public safety, but one has 385 00:38:02,380 --> 00:38:13,060 to wonder, did somebody just already have a vehicle set up with four cameras? 386 00:38:13,060 --> 00:38:18,820 And they say, oh, hey, don't you think it would be a good idea if everybody had four 387 00:38:18,820 --> 00:38:21,440 cameras on their vehicle? 388 00:38:21,440 --> 00:38:25,700 And then the regulator says, oh, yeah, that does sound like a good idea. 389 00:38:25,700 --> 00:38:30,700 It sounds like that would be safe and for the best of everybody. 390 00:38:30,700 --> 00:38:37,420 Now they pass a law that says, OK, all vehicles have to have four cameras on them. 391 00:38:37,420 --> 00:38:44,860 Well now if you're starting a delivery company, that's essentially going to be a cost to entry. 392 00:38:44,860 --> 00:38:51,140 You're going to have to install four cameras on your vehicle. 393 00:38:51,140 --> 00:38:58,700 So that may not be the best example, but you may be able to think of better ones. 394 00:38:58,700 --> 00:39:06,220 And so by all means, brainstorm for better or for worse. 395 00:39:06,220 --> 00:39:17,980 But that's sort of the idea is, producers will advocate for regulations that they think 396 00:39:17,980 --> 00:39:24,580 they can jump through, so they think they can meet that regulatory requirement. 397 00:39:24,580 --> 00:39:32,740 But that regulatory requirement may box out competition. 398 00:39:32,740 --> 00:39:35,460 And you don't have to take my word for it. 399 00:39:35,460 --> 00:39:43,300 So a lot of times if you go to, so I've attended a handful of cannabis conferences and you'll 400 00:39:43,300 --> 00:39:49,980 talk to your seminars, and people will be pretty upfront. 401 00:39:49,980 --> 00:39:57,940 Producers will be quite upfront that they recommend going in, sitting in on the agency 402 00:39:57,940 --> 00:40:07,220 meetings and being the squeaky wheel in these regulators' ears. 403 00:40:07,220 --> 00:40:10,740 And I mean, one almost can't necessarily fault them. 404 00:40:10,740 --> 00:40:19,180 I mean, they're parties in the industry and if they're trying to maximize their profits 405 00:40:19,180 --> 00:40:26,860 and that's how they can do so, then that's how they're going to act. 406 00:40:26,860 --> 00:40:30,340 So I think it just should be something kind of taken into consideration. 407 00:40:30,340 --> 00:40:36,740 And long story short is, I think consumers should be at the table, right? 408 00:40:36,740 --> 00:40:43,780 Because I heard once, if you're not at the table, you're for lunch. 409 00:40:43,780 --> 00:40:57,380 So if they have these meetings and it's just the regulatory agency, their representatives, 410 00:40:57,380 --> 00:41:01,660 and then you just have the producers. 411 00:41:01,660 --> 00:41:07,860 Well the consumers aren't there at the table, right, and the consumers have a stake in the 412 00:41:07,860 --> 00:41:09,900 market. 413 00:41:09,900 --> 00:41:14,700 So they may be getting the short end of the stick, right? 414 00:41:14,700 --> 00:41:30,100 And so I think that's ultimately up to the consumers to get to the table and to push 415 00:41:30,100 --> 00:41:34,100 forth what they may wish. 416 00:41:34,100 --> 00:41:36,620 Or for other producers, right? 417 00:41:36,620 --> 00:41:43,380 If you're a smaller producer, then you may want to go to the board meetings and whatnot 418 00:41:43,380 --> 00:41:46,100 and speak up for yourself too. 419 00:41:46,100 --> 00:41:51,380 Because like I said, the large companies, you can count on it that they're going to 420 00:41:51,380 --> 00:42:00,140 be at the meetings and they're going to be advocating for policies that may, they may 421 00:42:00,140 --> 00:42:05,420 on their face seem like they may be beneficial to the public. 422 00:42:05,420 --> 00:42:18,660 However, they may be cleverly designed to box out competition. 423 00:42:18,660 --> 00:42:30,220 So I'm not going to hammer that home too much further, but something that's taken into consideration. 424 00:42:30,220 --> 00:42:38,460 Well we talked about the economic theory, and I think it's kind of, it can only get 425 00:42:38,460 --> 00:42:39,460 us so far. 426 00:42:39,460 --> 00:42:43,140 You can kind of game theory it out so much. 427 00:42:43,140 --> 00:42:45,940 What if player A does X? 428 00:42:45,940 --> 00:42:50,380 What if player Y does Y? 429 00:42:50,380 --> 00:42:53,420 And so on and so forth. 430 00:42:53,420 --> 00:42:56,540 But at a certain point, you actually have to look at the empirics, right? 431 00:42:56,540 --> 00:42:59,620 You actually have to look at how many firms are in the market. 432 00:42:59,620 --> 00:43:01,340 What's the output? 433 00:43:01,340 --> 00:43:02,340 What are the prices? 434 00:43:02,340 --> 00:43:08,060 And that's what's called performance. 435 00:43:08,060 --> 00:43:14,980 So how are we going to go about measuring market performance? 436 00:43:14,980 --> 00:43:22,580 Well, we can look at profitability, how profitable are the firms. 437 00:43:22,580 --> 00:43:31,140 This is tricky, but we've kind of shown that we can almost tackle this. 438 00:43:31,140 --> 00:43:35,180 So we can make an estimated profitability. 439 00:43:35,180 --> 00:43:43,260 It would be awesome to actually have everyone's balance sheets and actually have an actual 440 00:43:43,260 --> 00:43:47,380 measure of everyone's profit or loss. 441 00:43:47,380 --> 00:43:54,420 That would be cool to do really rich statistics, but no one's going to just turn over their 442 00:43:54,420 --> 00:43:56,260 balance sheets like that. 443 00:43:56,260 --> 00:43:59,000 So we can estimate it. 444 00:43:59,000 --> 00:44:02,740 So we can look at the revenues. 445 00:44:02,740 --> 00:44:09,540 We can look at average revenue, and we can try to estimate the prices of inputs, and 446 00:44:09,540 --> 00:44:13,500 we can estimate profitability. 447 00:44:13,500 --> 00:44:18,260 Once again, with concentration, the state of Massachusetts could look at this because 448 00:44:18,260 --> 00:44:25,060 they probably have the sales per retailer and so on and so forth. 449 00:44:25,060 --> 00:44:35,860 So they could see if the market's becoming more concentrated or not over time or less 450 00:44:35,860 --> 00:44:38,540 concentrated over time. 451 00:44:38,540 --> 00:44:48,380 We can just look at the number of players in the market, and we can even see if average 452 00:44:48,380 --> 00:44:51,940 profits are going up or down. 453 00:44:51,940 --> 00:44:59,020 So if average profits are going up, well, the market may be getting a bit more concentrated, 454 00:44:59,020 --> 00:45:10,380 and these players may be finding a way to squeak price up and up and up above cost. 455 00:45:10,380 --> 00:45:16,700 If average profits are going down, the market may be becoming more concentrated over time, 456 00:45:16,700 --> 00:45:24,860 I mean less concentrated over time, and price is falling closer and closer and closer to 457 00:45:24,860 --> 00:45:27,700 average cost. 458 00:45:27,700 --> 00:45:37,500 Because that's sort of the theory of economics is given perfect competition, price will fall 459 00:45:37,500 --> 00:45:39,500 to average cost. 460 00:45:39,500 --> 00:45:44,220 However, we don't ever have quite perfect competition, right? 461 00:45:44,220 --> 00:45:54,580 Because we don't ever have perfect information, or there's structural barriers to entry, or 462 00:45:54,580 --> 00:45:59,340 there's transaction costs, and so on and so forth. 463 00:45:59,340 --> 00:46:02,180 So all of that brings us to our next bullet point. 464 00:46:02,180 --> 00:46:07,460 Well, it's useful to measure these barriers to entry. 465 00:46:07,460 --> 00:46:11,940 And this is a bullet point I wanted to say is understudied. 466 00:46:11,940 --> 00:46:17,080 So how do you even go about measuring barriers to entry? 467 00:46:17,080 --> 00:46:23,700 This is difficult, can be arbitrary, yet valuable. 468 00:46:23,700 --> 00:46:32,820 So the only idea I've had so far is maybe you could look at license fees by state and 469 00:46:32,820 --> 00:46:40,540 see if that affects the structure, the performance in any way. 470 00:46:40,540 --> 00:46:46,700 But yet again, it's really hard to measure, like the license fees, it's hard to get a 471 00:46:46,700 --> 00:46:48,620 gauge of, right? 472 00:46:48,620 --> 00:46:52,580 So for example, how do you compare apples to apples? 473 00:46:52,580 --> 00:47:02,140 So the different states have different licensing structures, different fee structures. 474 00:47:02,140 --> 00:47:05,300 How do you measure the difficulty of an application? 475 00:47:05,300 --> 00:47:13,780 What if one application takes you six months, and the other one takes you a year? 476 00:47:13,780 --> 00:47:20,060 How do you go about measuring some of these barriers to entry? 477 00:47:20,060 --> 00:47:26,740 Is it worthwhile if you're up to the challenge? 478 00:47:26,740 --> 00:47:31,860 And then finally, there's this measure of the total factor of productivity. 479 00:47:31,860 --> 00:47:38,260 And so this is something, a measure that was put forth by Joseph Stigler. 480 00:47:38,260 --> 00:47:40,820 He was at least one who utilized it. 481 00:47:40,820 --> 00:47:43,620 And that's essentially what we've been estimating. 482 00:47:43,620 --> 00:47:50,700 And that's just centrally sales per labor and sales per capital. 483 00:47:50,700 --> 00:47:54,660 And so we've been getting at those measures. 484 00:47:54,660 --> 00:48:07,940 And we actually kind of nearing the end, but just in the last bit of time here, let's get 485 00:48:07,940 --> 00:48:10,140 at those measures. 486 00:48:10,140 --> 00:48:18,340 So for those of you that are just joining us, we've been working with this data from 487 00:48:18,340 --> 00:48:20,740 Massachusetts. 488 00:48:20,740 --> 00:48:29,940 You can find the data through the Socrata API. 489 00:48:29,940 --> 00:48:36,220 And please tune into prior episodes or Saturday morning statistics. 490 00:48:36,220 --> 00:48:43,820 And I can help you or email me or what have you, and can help walk you through this. 491 00:48:43,820 --> 00:48:51,820 But just since we only have a limited amount of time, long story short, we read in the 492 00:48:51,820 --> 00:48:57,260 data here from Massachusetts. 493 00:48:57,260 --> 00:49:10,460 We create forecasts using an ARIMA model. 494 00:49:10,460 --> 00:49:14,180 And we forecast sales. 495 00:49:14,180 --> 00:49:28,540 We forecast plants, employees, retailers, cultivators, and total licensees. 496 00:49:28,540 --> 00:49:40,180 And that allows us to predict sales per retailer, plants per cultivator, and employees per licensee. 497 00:49:40,180 --> 00:49:52,540 And so these are the forecasts that we have plotted that I showed you at the beginning. 498 00:49:52,540 --> 00:50:01,860 And then, oh yes, the total factor of productivity. 499 00:50:01,860 --> 00:50:09,780 And so this is where, remember I showed you the Cobb-Douglas production function in prior 500 00:50:09,780 --> 00:50:10,860 weeks. 501 00:50:10,860 --> 00:50:17,180 So here we're essentially estimating a Cobb-Douglas production function for the cannabis industry, 502 00:50:17,180 --> 00:50:25,860 where we've got output, y, so this is going to be our sales, times technology. 503 00:50:25,860 --> 00:50:30,540 And I want to stress this point real quick. 504 00:50:30,540 --> 00:50:32,900 Technology we're taking as a constant. 505 00:50:32,900 --> 00:50:35,500 Well, guess what? 506 00:50:35,500 --> 00:50:39,340 Technology is not constant. 507 00:50:39,340 --> 00:50:44,980 And this is where you can do interesting analysis with your production function. 508 00:50:44,980 --> 00:50:51,780 So one of my favorites is just saying you can create what's called regime models. 509 00:50:51,780 --> 00:50:54,980 So A0 and A1. 510 00:50:54,980 --> 00:51:03,700 So you could say, OK, what if technology switches between two different regimes? 511 00:51:03,700 --> 00:51:16,180 And so that's where we could say, oh, maybe we had one technology here from 2019 to 2020. 512 00:51:16,180 --> 00:51:25,100 And we may have had a different technology regime going on here in 2021. 513 00:51:25,100 --> 00:51:35,260 And technology is sort of this abstract word that economists use to just capture everything 514 00:51:35,260 --> 00:51:38,460 else besides from capital and labor. 515 00:51:38,460 --> 00:51:43,980 So this is sort of the state of the economy. 516 00:51:43,980 --> 00:51:50,060 So A is sort of a big deal. 517 00:51:50,060 --> 00:51:51,540 We're just going to take it as constant. 518 00:51:51,540 --> 00:51:57,940 But I just want to stress on the fact that if you let technology be dynamic, you can 519 00:51:57,940 --> 00:52:01,660 add a lot. 520 00:52:01,660 --> 00:52:05,980 You can make your analysis even that much more in-depth. 521 00:52:05,980 --> 00:52:09,780 But we're just going to take technology as constant. 522 00:52:09,780 --> 00:52:19,740 And we're just going to say, OK, labor and capital, which we are proxying as plants, 523 00:52:19,740 --> 00:52:29,020 is productive to a certain degree, alpha for the plants, beta for the labor. 524 00:52:29,020 --> 00:52:35,500 And with this production function, which we expect to be concave, so to have diminishing 525 00:52:35,500 --> 00:52:41,940 marginal returns, so the more and more we produce, the incremental amount is less and 526 00:52:41,940 --> 00:52:48,780 less, yet always increasing. 527 00:52:48,780 --> 00:52:55,260 So long story short, we expect alpha plus beta to be less than or equal to 1. 528 00:52:55,260 --> 00:53:00,980 And then empirically, they've measured alpha. 529 00:53:00,980 --> 00:53:01,980 There was somewhere here. 530 00:53:01,980 --> 00:53:04,580 But like I said, Wikipedia is not the best source. 531 00:53:04,580 --> 00:53:07,180 But it's just what I'm using for today. 532 00:53:07,180 --> 00:53:18,020 But historically, they were measuring, I think they were seeing alpha to be 0.3 and beta 533 00:53:18,020 --> 00:53:20,940 to be about 0.7. 534 00:53:20,940 --> 00:53:27,380 So real quick, in just a matter of minutes, let's see if we can measure some of these 535 00:53:27,380 --> 00:53:29,540 factors here. 536 00:53:29,540 --> 00:53:39,300 So we're going to basically just supplement the data here with weekly hours from the Federal 537 00:53:39,300 --> 00:53:40,300 Reserve. 538 00:53:40,300 --> 00:53:50,220 So this is just the average weekly hours of somebody in Massachusetts. 539 00:53:50,220 --> 00:53:55,340 So they're working around 33 hours per week. 540 00:53:55,340 --> 00:54:00,820 Then they're going up to around 34. 541 00:54:00,820 --> 00:54:09,020 And the reason we're doing this is because I'm using hours, total hours worked as a measure 542 00:54:09,020 --> 00:54:10,020 of labor. 543 00:54:10,020 --> 00:54:17,260 And the reason I'm doing that is because we can estimate the price of labor. 544 00:54:17,260 --> 00:54:22,580 And if we use hours, well, that's dollars per hour. 545 00:54:22,580 --> 00:54:27,820 And a lot of people get paid in dollars per hour. 546 00:54:27,820 --> 00:54:28,820 Your wage. 547 00:54:28,820 --> 00:54:34,160 So that's why we're estimating that. 548 00:54:34,160 --> 00:54:46,340 So long story short, you've got Y, K, and L. We're going to limit it here to we only 549 00:54:46,340 --> 00:54:51,220 have weekly hours through the end of August. 550 00:54:51,220 --> 00:55:03,020 And we're going to exclude the turbulent period. 551 00:55:03,020 --> 00:55:08,860 So actually, we're actually including the turbulent period. 552 00:55:08,860 --> 00:55:18,180 We're basically starting from when things open back up after May of 2020. 553 00:55:18,180 --> 00:55:30,020 So long story short, these are basically the total factors of productivity right here. 554 00:55:30,020 --> 00:55:33,740 So that's YPK. 555 00:55:33,740 --> 00:55:42,220 So you'll hear economists say, oh, YPL, KPL, or YPK and whatnot. 556 00:55:42,220 --> 00:55:47,140 So these are measures that economists like to look at. 557 00:55:47,140 --> 00:55:51,660 So here is YPL. 558 00:55:51,660 --> 00:55:54,980 So that output per labor. 559 00:55:54,980 --> 00:56:03,340 And so as you can see, labor is just incredibly productive right when the industry starts, 560 00:56:03,340 --> 00:56:05,540 which you would expect. 561 00:56:05,540 --> 00:56:11,620 Everyone's coming online, adding one more employee in those early days, which is really 562 00:56:11,620 --> 00:56:23,940 adding a lot. 563 00:56:23,940 --> 00:56:31,260 This is something that fluctuates, so something worth looking at. 564 00:56:31,260 --> 00:56:37,060 And then you can also look at YPK. 565 00:56:37,060 --> 00:56:43,900 That's essentially how productive your plants are going to be. 566 00:56:43,900 --> 00:56:57,720 Once again, your plants are super valuable right when the industry starts. 567 00:56:57,720 --> 00:57:05,340 If you look at the last 30 weeks, we've got a little bit of a negative trend here. 568 00:57:05,340 --> 00:57:08,900 Let's look at the last year. 569 00:57:08,900 --> 00:57:19,060 So here is this is basically sales per plant, so a crude measure of how productive your 570 00:57:19,060 --> 00:57:23,020 plants are. 571 00:57:23,020 --> 00:57:30,100 Well, let's estimate this Cobb-Douglas production function real quick. 572 00:57:30,100 --> 00:57:33,220 Here I did it in a fancy way. 573 00:57:33,220 --> 00:57:45,060 I'm going to not do it this way. 574 00:57:45,060 --> 00:57:51,500 Let's do the log. 575 00:57:51,500 --> 00:57:54,060 Let's do it this way. 576 00:57:54,060 --> 00:57:59,660 Bear with me real quick, and then we'll be wrapping up here since we're at the end. 577 00:57:59,660 --> 00:58:08,180 One second. 578 00:58:08,180 --> 00:58:19,260 I think one second. 579 00:58:19,260 --> 00:58:32,740 And I just have to figure out how to make a quick array here with a constant in it. 580 00:58:32,740 --> 00:58:43,140 We may have to save this for next time, but I would like to run the regression. 581 00:58:43,140 --> 00:58:52,500 Bear with me. 582 00:58:52,500 --> 00:58:59,700 Okay, unfortunately, hold on. 583 00:58:59,700 --> 00:59:04,500 Let's see if we can't do this real quick. 584 00:59:04,500 --> 00:59:13,340 Okay, if I can make this in one minute, then we'll do it, if not, I'll save it for next 585 00:59:13,340 --> 00:59:14,340 time. 586 00:59:14,340 --> 00:59:21,780 I apologize that this wasn't already set up here. 587 00:59:21,780 --> 00:59:27,380 Okay, I think I may be able to do this for you real quick. 588 00:59:27,380 --> 00:59:35,140 So let's just define our regressors here. 589 00:59:35,140 --> 00:59:55,500 Log A and log L. No promises, but we may be able to estimate alpha and beta real quick. 590 00:59:55,500 --> 01:00:13,220 Okay, so unfortunately. 591 01:00:13,220 --> 01:00:28,620 Unfortunately, I'm not able to code this up for you real quick on the fly. 592 01:00:28,620 --> 01:00:36,780 But that's okay, because we can pick this back up next week. 593 01:00:36,780 --> 01:00:42,980 Sometimes when I'm presenting, I just need to just sit down and think about the code 594 01:00:42,980 --> 01:00:49,540 a little bit, a little bit more to just kind of know what I'm actually doing. 595 01:00:49,540 --> 01:00:53,340 So my apologies, but I'm just going to need to code this up for you. 596 01:00:53,340 --> 01:00:56,460 But long story short, we'll pick up with this next week. 597 01:00:56,460 --> 01:01:01,860 And we're going to measure these total factors of productivity. 598 01:01:01,860 --> 01:01:08,860 So that way we can get our estimates of alpha and beta. 599 01:01:08,860 --> 01:01:17,940 And what's cool about measuring alpha and beta in the Cobb-Douglas production function 600 01:01:17,940 --> 01:01:28,700 is it will allow us to estimate the competitive wage rate, as well as the rate of return on 601 01:01:28,700 --> 01:01:31,220 capital. 602 01:01:31,220 --> 01:01:32,780 Why is that cool? 603 01:01:32,780 --> 01:01:40,460 Well, one, we can compare the competitive wage rate in the cannabis industry to other industries. 604 01:01:40,460 --> 01:01:41,460 Right. 605 01:01:41,460 --> 01:01:49,420 So you can look at the average wage of anything from say fast food to banking. 606 01:01:49,420 --> 01:01:56,580 So that way you can see where the cannabis industry workers fall out in the grand scheme 607 01:01:56,580 --> 01:01:58,080 of things. 608 01:01:58,080 --> 01:01:59,740 You can look at the interest rate. 609 01:01:59,740 --> 01:02:04,900 And this will be helpful for investors knowing, you know, should I invest in the cannabis 610 01:02:04,900 --> 01:02:09,300 industry and what rate of return should we expect? 611 01:02:09,300 --> 01:02:14,220 And it will let us measure or at least estimate profitability. 612 01:02:14,220 --> 01:02:15,220 Right. 613 01:02:15,220 --> 01:02:24,220 Because now if we have a measure of average sales, well, we can then estimate average 614 01:02:24,220 --> 01:02:37,300 cost and we'll guess what the difference between average sales and average cost is expected 615 01:02:37,300 --> 01:02:39,020 average profit. 616 01:02:39,020 --> 01:02:43,940 This isn't going to be the profit everybody's getting, but it's going to be what we would 617 01:02:43,940 --> 01:02:47,420 expect your average firm to make. 618 01:02:47,420 --> 01:02:50,780 And so then you can compare that to your own profit. 619 01:02:50,780 --> 01:02:57,100 So if you're a retailer in the space and we estimate average profits to be X, are you 620 01:02:57,100 --> 01:02:58,940 above or below? 621 01:02:58,940 --> 01:03:05,860 So then you can brag to your investors, hey, we're above expected profits or we're below 622 01:03:05,860 --> 01:03:07,140 expected profits. 623 01:03:07,140 --> 01:03:10,380 And you know, you wouldn't be bragging about that. 624 01:03:10,380 --> 01:03:17,860 You may be nudging, you may be telling your general manager that and trying to bark at 625 01:03:17,860 --> 01:03:23,380 them to to drum up sales a bit. 626 01:03:23,380 --> 01:03:25,660 So it's really helpful, right? 627 01:03:25,660 --> 01:03:31,340 Because you know, you're not operating in a silo here. 628 01:03:31,340 --> 01:03:34,900 You know, you've got other players in the market. 629 01:03:34,900 --> 01:03:41,180 And so as a retailer, you want to know how you size up against your competition. 630 01:03:41,180 --> 01:03:42,180 Same with the cultivators. 631 01:03:42,180 --> 01:03:45,300 You want to know how you size up. 632 01:03:45,300 --> 01:03:51,860 So that's what we'll work on next week is actually measuring next week. 633 01:03:51,860 --> 01:04:00,260 We'll estimate profitability by calculating these total factors of productivity. 634 01:04:00,260 --> 01:04:06,100 And then we can make statements about what we think about concentration and barriers 635 01:04:06,100 --> 01:04:08,860 to entry. 636 01:04:08,860 --> 01:04:17,860 So stay tuned for next week, and I will iron out the estimation of these parameters for 637 01:04:17,860 --> 01:04:18,860 us. 638 01:04:18,860 --> 01:04:24,140 My apologies that I couldn't do it on the fly for today, but that's probably for the 639 01:04:24,140 --> 01:04:32,660 best, because that way we can do it proper and thoroughly next week. 640 01:04:32,660 --> 01:04:45,660 And so next week, we'll add at least two new series and forecasts will add the competitive 641 01:04:45,660 --> 01:04:53,440 historic wage rate and the competitive interest rate for plants. 642 01:04:53,440 --> 01:05:01,460 So you can know what rate of return on average having one plant in the ground is. 643 01:05:01,460 --> 01:05:06,140 So that's where we'll pick up next week. 644 01:05:06,140 --> 01:05:12,580 Want to go ahead and thank everybody for attending to rain a little extra today. 645 01:05:12,580 --> 01:05:25,500 So any questions, comments, concerns, ideas from anyone from the group? 646 01:05:25,500 --> 01:05:27,460 On that case. 647 01:05:27,460 --> 01:05:29,140 I'm always here. 648 01:05:29,140 --> 01:05:30,140 Send me a message. 649 01:05:30,140 --> 01:05:32,260 Send me an email, get in contact with Canelitics. 650 01:05:32,260 --> 01:05:34,900 We're always here to help the industry. 651 01:05:34,900 --> 01:05:41,580 So until next week, keep your nose to the grindstone. 652 01:05:41,580 --> 01:05:44,200 Stay productive. 653 01:05:44,200 --> 01:05:47,460 Have fun and enjoy yourselves. 654 01:05:47,460 --> 01:05:50,300 Thanks, Keegan. 655 01:05:50,300 --> 01:05:52,940 Definitely, definitely, definitely. 656 01:05:52,940 --> 01:05:55,940 Thank you very much. 657 01:05:55,940 --> 01:06:06,740 Bye now.