PropertyChat 2018 stats

You see, this is where this idea is completely problematic.

You assert that your post is "pointless", whereas I see it as several completely valid questions - "how do we measure value?" and then, "how do we measure the change in value added over time?".

The question of value is so incredibly subjective that it's almost impossible to measure.

Even a bunch of silly threads in the coffee lounge adds value to the community in their own way.

Alas, I fear that any objective measure of "value" is likely to elude us and we just have to make do with the simplistic data that is easy to generate. Then we can just sit around and argue about exactly what that data actually means!

 

9. Threads per day in 2018 (28 day moving average)



I think that's much "flatter" than what we see in posts per day - far less fluctuation / more consistent (between the beginning of April and the end of November, we ranged between 17 and 20 new threads per day on average).

Comparing to previous years:

Threads per day 2016-18



Surprisingly we saw a bit more volatility in 2017 - but overall I think they are much "closer" between each year.

 

10. Threads per month in 2018

 

11. Threads per year

 

12. Active users per day in 2018 (Number of members who have logged in each day - 28 day moving average)



You can really see the Christmas dip!

Active users per day 2016-18



What on earth happened in July 2016 ??

• 01 Jul - 701
• 02 Jul - 665
• 03 Jul - 553
• 04 Jul - 618
• 05 Jul - 724
• 06 Jul - 717
• 07 Jul - 645
• 08 Jul - 499
• 09 Jul - 474
• 10 Jul - 433
• 11 Jul - 498
• 12 Jul - 571
• 13 Jul - 605
• 14 Jul - 566
• 15 Jul - 571
• 16 Jul - 534
• 17 Jul - 454
• 18 Jul - 500
• 19 Jul - 577
• 20 Jul - 606
• 21 Jul - 600
• 22 Jul - 587
• 23 Jul - 560
• 24 Jul - 483
• 25 Jul - 573
• 26 Jul - 623
• 27 Jul - 645
• 28 Jul - 633
• 29 Jul - 634
• 30 Jul - 594
• 31 Jul - 497

The figures in bold are way below average. I'm wondering if there was something happening at the server end of the website because that is way below the average and such sharp drops haven't happened at any other time

The way this figure is calculated is not something I can recalculate - it's calculated daily by the software and stored, so there is scope for anomalies to creep in there.

Anyway - as you can see we actually had more daily active users in 2018 than in 2016, but not quite as many as in 2017.

Unfortunately, we can't generate meaningful stats for active users per month, since it's a figure stored by the database and not something I can recalculate. Since it's counting unique active members, simply adding up how many daily active users there were for the month is meaningless, since people generally log on more than once per month.

... but we can still look at users who post instead!

 

Absolutely.

There was a definite change about the type of threads being posted around the time that APRA effectively put the breaks on the property market. Combined with the banking royal commission - the lending landscape changed dramatically and I think we're seeing this reflected in forum activity.

Post volume actually went up for a while as people sought to come to terms with what that all meant - but then it dropped off quite a bit since there were a lot of people who decided to sit on the sidelines.

I'll try and do some reports which look at which topics the posts came from.

It would be useful if any of the brokers could go back through the finance archives and identify major announcements from APRA that were posted about on here that we can use to correlate with the post volume data.

 

Okay - here is some data which pretty comprehensively shows what I've been saying.

I looked at posts per day since launch for our 3 biggest topics:

• Where to buy
• Property Market Economics
• Finance

The chart below shows 90 day moving averages of daily post rates made in those topics:



You can clearly see the sharp decline in post volumes in the Where to Buy forum since July 2017.

Similarly, there was an immediate sharp drop in finance discussions from July 2017 through to October 2017 at which point the discussion levelled off, but still maintained a definite downtrend.

At the same time, there was a rapid increase in the volume of Property Market Economics posts from April - July 2017, with a similar peak in July 2017 followed by a sharp decline through to October 2017, then more peaks and troughs - but a definite uptrend over time to today where we now get significantly more posts in the economics forum than we do in either Where to Buy or Finance.

I think that's a pretty clear picture of investor sentiment there.

I'd be interested to see the dates of major announcements about APRA changes and how they correlate here - noting that the above chart uses 90 day moving averages and so the reaction to the announcements in the chart may be delayed by up to 3 months.

 

Give me a list of keywords I can add to my queries:

Positive:

• boom
• bubble (is this positive or negative? - I'm going to leave it in positive because it is usually used during a growing market)

Negative:

• bust
• correction
• crash

What other words can I use to indicate sentiment?

 

Here is a chart showing usage of the word "APRA" in posts since we launched.

Chart shows how many posts per day have used the term APRA.



Note that this is showing the 90 day moving average - so data may lag by up to 3 months.

 

@Terry_w total posts: 19,515 ... PropertyChat total posts: 646,497 ... percentage: 3.0% of total

 

I found the previous chart about posts mentioning APRA curious in that we started with a very high post rate back after launch of the site - so I decided to check out the same data from Somersoft to see if there was an earlier trend.

Here is the data from Somersoft showing the average posts per day which mention the word "APRA" (90 day moving average):



Almost nothing until 2014 - at which point it starts to pick up ... but that spike at the end!!!???

I decided to zoom in and look at a 28 day moving average from the beginning of January 2014 to see if we can see a more definite pattern:



All this APRA stuff really kicked off in May 2015!!

It's no wonder we were getting such high post counts with the APRA keyword right after the launch of PropertyChat in June 2015!

I looked through some of the threads that were getting traffic back in May 2015 - here's some of the discussions we were having back then:

• RBA May 2015 announcement - 25 point cut
  
• Finance risk while growing a large portfolio
  
• Changes / tightening on servicing for investors
  
• What sort of impact will all this tightening of serviceability have on the market?

 

Okay then, so using @wombat777 's suggestion of keywords - here is a chart showing the percentage of posts which contain "positive" keywords vs those which contain "negative" keywords. We can use this as a crude form of investor sentiment.



Note that this is across all forum topics.

The crossover point where the "sentiment" changes to being predominantly negative was just before the middle of November, 2018.

Let's look at the difference between the two values charted above:

 

Here's the same queries run on Somersoft (2002 - 2015)





Surprisingly, there was only one period where sentiment dropped into the negative - between August and October 2011!?

 

Okay, back to our regularly scheduled programming ...

13. Posting users per day in 2018 (Number of members who have posted at least once each day - 28 day moving average)



Posting users per day 2016-18