Some fun with the SNDB - Data is beautiful

News: Welcome to the Canadian Paper Money Forum.

So I have been playing a bit with the SNDB recently. There was some intel (min, max, population, plate numbers analysis/research that I don't understand yet...), but with dates and locations, there was probably a way to tell some stories.. And in short : definitely !

Some context :

• The data is limited to 10$ polymers, andwas extracted prefix by prefix with the query thingy. Not tedious, but it was merely to test the data and I didn't want to generate too much traffic
• To update the graph, I'd need to re-query them
• There is ~130k banknotes in there
• Adding the signatures/splitting the changeovers was done manually (easy to automate if need be
• Data was kept raw, except maybe trials to remove sequential notes - not worth it, and tedious in Excel
• The rest is easy Excel stuff, but the computer is working a bit !

Some high-level storytelling (in no order) :

• There doesn't seem to be a precise seasonality for peak submissions in the DB, but there are clear lows June to August
• There is less data as of 2024, and more data the better, but the trend information still seems useable. I have not checked yet for geographical representativeness (which could be more of an issue)
• It took 18 months for the Wilkins-Macklem prefixes to see the light of day, but only 13 for the Rogers-Macklem ones (there was of course the Lane hiatus - we could derive cross denomination intel by analyzing the 5$ data
• Some prefixes started slow but went full speed after some times (topic of a future post...)
• Retiring the MacDonald horizontals was surprisingly quick - around two month - and following a push in the Vertical series
• My initial opinion on sequential numbers : it's best to have the data in the SNDB. It may be useful for use cases (it's the only way to identify what are likely newly circulating banknotes, which may be useful for some usecases (eg. geographical distribution of new banknotes origination, per prefix). Easy to remove sequential numbers (or sample/keep a given percentage) with proper tools if the use case requires less outliers. 
• Further analysis of the MacDonald's retirement could yield some intel as to how it could play out with the introduction of the Vertical 5$ next year.

I parsed and pivoted the data in a few ways and will post them soon. Meanwhile - suggestions, comments, ideas, use cases are definitely welcome.

It's easy to have fun with clean data. 

Wow I found this very interesting and thank you for posting it on the CPM forum but I must ask. How long did it take you to create this SNDB $10 data? I was surprised there was 130k banknotes (polymer series that was entered over the years). Also what does the colours mean inside the photo? I saw red-yellow-green range for the $10 prefixes that were coloured in by each months of the year depending on the series/prefix?

I try my best to enter every banknote I get into the SNDB and I even go to the bank to withdraw large sums of cash to search and enter it in the SNDB. The more people enters into the SNDB, the data gets better.

Also do you plan to do $5, $20, $50 and $100 denomination in the future? I might do some data research about the $10 prefix when I have time for the future.

I was inspired by this post and I decided to see how many notes were entered for the 2013 Series prefix (FEW) first ever frontier prefix produced into circulation. (This took me a long time, I would've done all of the $10 prefix in the SNDB from FEW-FTV (2013 series), CDA-CDF (2017 Series), and FTW-present (2018 Series) but I don't have time for it.)

Enjoy the data!

2013-11   82
2013-12   73
2014-01   65
2014-02   54
2014-03   80
2014-04   80
2014-05   120
2014-06   89
2014-07   91
2014-08   91
2014-09   156
2014-10   179
2014-11   183
2014-12   152
2015-01   110
2015-02   69
2015-03   104
2015-04   52
2015-05   27
2015-06   58
2015-07   71
2015-08   67
2015-09   58
2015-10   128
2015-11   134
2015-12   161
2016-01   166
2016-02   115
2016-03   150
2016-04   56
2016-05   70
2016-06   37
2016-07   35
2016-08   58
2016-09   61
2016-10   379
2016-11   82
2016-12   45
2017-01   46
2017-02   53
2017-03   46
2017-04   45
2017-05   38
2017-06   20
2017-07   21
2017-08   53
2017-09   89
2017-10   47
2017-11   82
2017-12   34
2018-01   92
2018-02   28
2018-03   56
2018-04   32
2018-05   2
2018-06   1
2018-07   0
2018-08   0
2018-09   0
2018-10   3
2018-11   22
2018-12   36
2019-01   26
2019-02   21
2019-03   35
2019-04   24
2019-05   29
2019-06   11
2019-07   6
2019-08   14
2019-09   13
2019-10   10
2019-11   13
2019-12   23
2020-01   21
2020-02   17
2020-03   13
2020-04   22
2020-05   13
2020-06   18
2020-07   5
2020-08   20
2020-09   19
2020-10   41
2020-11   17
2020-12   33
2021-01   20
2021-02   5
2021-03   8
2021-04   5
2021-05   19
2021-06   11
2021-07   18
2021-08   18
2021-09   8
2021-10   11
2021-11   13
2021-12   3
2022-01   4
2022-02   29
2022-03   41
2022-04   8
2022-05   5
2022-06   7
2022-07   4
2022-08   6
2022-09   24
2022-10   9
2022-11   7
2022-12   19
2023-01   13
2023-02   22
2023-03   4
2023-04   0
2023-05   3
2023-06   6
2023-07   1
2023-08   0
2023-09   0
2023-10   1
2023-11   0
2023-12   1
2024-01   5
2024-02   0
2024-03   0
2024-04   0
2024-05   1
2024-06   1
2024-07   4
2024-08   1
2024-09   2
2024-10   1
2024-11   2
2024-12   0
2025-01   2
2025-02   1
2025-03   2
2025-04   0
2025-05   0
2025-06   0
2025-07   1
2025-08   0
2025-09   0
2025-10   0
2025-11   0
2025-12   0
2026-01   1
2026-02   0

Total: 5475 notes as of Feb 2026

It seems like FEW these days is literally very few in circulation now.

AJG : I'm not there yet, but will let you know when I get to BIG ! 

TN56 : I don't recall it being that long. Changing the URL for the query, copy+paste in Excel, for around 50 prefixes, is fairly rhythmic. Probably more so than entering 100 notes in the DB. It took more time testing and adding columns (eg. calculating first date of appearance for any given prefix and calculating the number of months between a given entry and the first appearance of the prefix - even though fairly recent, my computer didn't enjoy that work on 140k lines).

Also, the issue is that one would have to re-download all the prefixes each month because it's not (yet?) possible to extract otherwise. There are ways to automate, but it can create a burden for websites if done incorrectly/some website owners have measures against that/some site owners do not allow or like automated download. So, yes, I'd like to extend the scope of the experience, but I'm still testing and discovering use cases with the 10$, the size is manageable and I finally have a project to learn SQL !

But the fun part is that flipping/pivoting/transforming the data afterwards is a matter of a couple minutes : have a look below at the FEW prefix - we have the same numbers

With regards to the color schemes, there are three heatmaps, where the number in the cell drives the cell's background color. I masked the numbers in the first post for aesthetic reasons (numbers were not necessary for that storytelling), but here's an extract of what it looks like with the numbers. The two other heatmaps (blue to red) are for lines (prefixes) and columns (months) totals.

Actually, why not go BIG now :

It took 18 months for the Wilkins-Macklem prefixes to see the light of day, but only 13 for the Rogers-Macklem ones (there was of course the Lane hiatus - we could derive cross denomination intel by analyzing the 5$ data

For the Rogers/Macklem $10s to take just 13 months to see the light of day nationwide as opposed to 18 months for the Wilkins/Macklem $10s, likely strengthens the possibility that demand for $10 bills have rebounded - at least in some provinces (including Newfoundland and Labrador), due to, of course, hefty inflation.  Plus, I did learn that inflation in Newfoundland and Labrador is higher than in Ontario - in fact, according to an article from some days ago, Newfoundland and Labrador's inflation rate is reported to be 2.4% (fifth highest in rank among the ten provinces). Three of the top five provinces with the highest inflation rates are all in Atlantic Canada- so that could explain the lack of new $5 notes in Atlantic Canada for some years now.  Ontario currently has an inflation rate of 2.0% - they are still seeing new $5 bills to this day, but once it gets to where Newfoundland and Labrador currently is, it will likely be their turn to see an increase in $10 bills in circulation.

And for the record, I did not see any fresh Wilkins-Macklem $10s in Newfoundland until late last summer or September after many FFNs the bank had in their inventory were exhausted, and the prefix was FFH.  It seems as if the $10 bill is likely the only banknote denomination that is rising in demand in my province (the other four are likely in decline due to rising debit card usage), and many of the fresh $10s being sent to Newfoundland and Labrador (and maybe Atlantic Canada in general) are replacing the $5 notes that are unfit for circulation.  I strongly doubt I am going to see any fresh $5 bills in my province until the Terry Fox version comes out - and even then, the $5 notes may become few and far between like the $10 bill had previously been for over three decades, and maybe the future lack of $5 bills in circulation could be worse than the $10 bill was.

I also learned that the vast majority of new $5s that are in circulation are in Ontario and Quebec - very few reported from BC and any of the Prairie provinces.  And of the IO* series, they were likely all in Ontario.  Quebec's inflation rate is the highest at 3.0%, so it's possible that demand for $5 bills may have started to decline there not that long ago.

For the Rogers/Macklem $10s to take just 13 months to see the light of day nationwide as opposed to 18 months for the Wilkins/Macklem $10s, likely strengthens the possibility that demand for $10 bills have rebounded - at least in some provinces (including Newfoundland and Labrador), due to, of course, hefty inflation.  Plus, I did learn that inflation in Newfoundland and Labrador is higher than in Ontario - in fact, according to an article from some days ago, Newfoundland and Labrador's inflation rate is reported to be 2.4% (fifth highest in rank among the ten provinces). Three of the top five provinces with the highest inflation rates are all in Atlantic Canada- so that could explain the lack of new $5 notes in Atlantic Canada for some years now.  Ontario currently has an inflation rate of 2.0% - they are still seeing new $5 bills to this day, but once it gets to where Newfoundland and Labrador currently is, it will likely be their turn to see an increase in $10 bills in circulation.

And for the record, I did not see any fresh Wilkins-Macklem $10s in Newfoundland until late last summer or September after many FFNs the bank had in their inventory were exhausted, and the prefix was FFH.  It seems as if the $10 bill is likely the only banknote denomination that is rising in demand in my province (the other four are likely in decline due to rising debit card usage), and many of the fresh $10s being sent to Newfoundland and Labrador (and maybe Atlantic Canada in general) are replacing the $5 notes that are unfit for circulation.  I strongly doubt I am going to see any fresh $5 bills in my province until the Terry Fox version comes out - and even then, the $5 notes may become few and far between like the $10 bill had previously been for over three decades, and maybe the future lack of $5 bills in circulation could be worse than the $10 bill was.

I also learned that the vast majority of new $5s that are in circulation are in Ontario and Quebec - very few reported from BC and any of the Prairie provinces.  And of the IO* series, they were likely all in Ontario.  Quebec's inflation rate is the highest at 3.0%, so it's possible that demand for $5 bills may have started to decline there not that long ago.

@AJG:

Here is a very interesting study by the Bank of Canada on Cash Use and Distribution that could provide answers to all of your questions.  It is a fascinating document.

https://www.bankofcanada.ca/wp-content/uploads/2025/08/sdp2025-12.pdf

Interestingly, there is a graph showing the condition of $50 and $5 notes "held" (in hand at that moment) or "last held". by survey participants.  For the $5s, the frequency of "held" notes in poor condition was 0.8% versus 2.5% of "last held" $5 notes.

A large percentage of $5 notes fell into the Good-Fair-Poor categories.