RTB Rent Register

[Banquo]

There is a difference between the register being inadequate for its purpose versus the views by some that the RTB are using it to deliberately shaft landlords.

My only feeling is that any such product will have technical issues in it when launched but overtime the data and product will improve. It's an opinion not a fact.

 

[stormy]

There is a difference between the register being inadequate for its purpose versus the views by some that the RTB are using it to deliberately shaft landlords.

Sorry, not sure I understand. The RTB have published a register which is not of any use in the determination of market rent. The majority of the entries in this register are below market rent (this is a fact and it is by construct). This register must be considered when setting or reviewing rents.

How is that not shafting accomodation providers?

 

[Greenbook]

This is especially the case as what the Rent Register is producing is to a large extent rent controlled and/or older rentals which will have rents on the low side.

Now it can't help producing rent controlled properties, but these are not market rents. Yet we still have to use them. And per the RTB anyway the core tool for determining market rent is the Register.

 

[patfert1]

Isn't the main issue with the register that it's not market rent?

@patfert1 stating the available data is a normal distribution is not the same as saying this is of use for setting market rent.

Right, two different things entirely.

Regarding market rent and the RTB register...

A "supply and demand" driven market rent for one specific property and one specific tenant, cannot be definition be represented by any register of rent amounts for different properties and different tenants. Plus, the RTB search query is relatively crude in that it does not take into account condition of the property, location advantages unique to the property, etc.

According to the legislation (from Google AI, I believe it is accurate):

Under the Residential Tenancies (Miscellaneous Provisions) Act 2026, effective from 1 March 2026, market rent is defined as the amount a tenant would reasonably pay for a similar home in a comparable area at the time the tenancy begins, determined by using the RTB rent register. Landlords must justify new rents based on comparable properties (similar BER, floor area, and bedrooms) and provide this evidence to the tenant and RTB.

These two things are not really compatible. I mean "the amount a tenant would reasonably pay for a similar home in a comparable area at the time the tenancy begins" suggests normal rules of supply and demand, whereas "determined by using the RTB rent register" refers to other properties which may be under RPZ, rents set years ago and not updated, new tenancies 24 months ago, etc.

Having read the RTB rent register algorithm kindly obtained by @Sr. Tayto, at first read I don't find anything particularly objectionable or blatantly biased in the algorithm, as I noted earlier. I could be wrong, but this is my working hypothesis. But since I could be wrong, I'd like to challenge this hypothesis. What I would like to do is the following:
1) Scrape a new set of very comprehensive data for a local electoral area (and later other LEAs). It should include a wide range of all the search criteria and represent all of the EDs in the LEA. I have the tool for this.
2) Implement the RTB algorithm and run it on this data set. First, see if I can get in-or-around the same 10 matches for a reference property, as the RTB register query tool. If yes, then...
3) Generate the match score for all of the other properties in the data set. Examine this data set to see if there are some or many which arguably should be in the top 10 matches, e.g. higher rent levels not in the original 10 which were left out due to the RTB algorithm, but which under a slightly different algorithm might have appeared. To be honest though, even if I find that for one reference property, it doesn't mean much unless I can get the same result for many other references properties, and in other parts of Ireland too. And I'm unlikely to have bandwidth for such a task. But anyway this is what I am working toward.

I'm working on item 2...

Thoughts and suggestions welcome!

 

[patfert1]

The RTB have published a register which is not of any use in the determination of market rent.

Just to add, I agree, but that is what the legislation asks the RTB to do, and trying not to be biased against the RTB without evidence, the tool they have produced appears to reasonably fairly do it's job (my first impression... see prev post). It's not a "market rent" though, I agree with you 110%.

 

[Banquo]

Just to add, I agree, but that is what the legislation asks the RTB to do, and trying not to be biased against the RTB without evidence, the tool they have produced appears to reasonably fairly do it's job (my first impression... see prev post).

This is exactly my point.

 

[stormy]

Just to add, I agree, but that is what the legislation asks the RTB to do

This is not correct. The legislation did not state just 10 tenancies would be returned when the register is checked.

 

[patfert1]

The legislation did not state just 10 tenancies would be returned when the register is checked

Yes you are spot on with that. Its a very arbitrary and small cut off. Why not show top 30, 50, whatever.

 

[stormy]

There is a difference between the register being inadequate for its purpose versus the views by some that the RTB are using it to deliberately shaft landlords.

Its a very arbitrary and small cut off. Why not show top 30, 50, whatever

By restricting the public extract to just 10 units, the odds of finding a recent tenancy that actually reflects current market rent drop massively. The sample size skews the data. That's a mathematical truism and the RTB is choosing to do this. The legislation simply says:
"(3) The extent of that extract shall be such as the Board determines is likely to make the published register useful to members of the public."

It’s an RTB policy to cap it at 10, not the law, and it deliberately skews the data to show artificially low rents. The obvious solution is to publish all rentals on the extract so providers can use the most recent comparables. By hiding the vast majority of the data, the RTB is completely shafting accommodation providers and pushing a false narrative about market rents.

 

[patfert1]

It’s an RTB policy to cap it at 10, not the law, and it deliberately skews the data to show artificially low rents. The obvious solution is to publish all rentals on the extract so providers can use the most recent comparables. By hiding the vast majority of the data, the RTB is completely shafting accommodation providers and pushing a false narrative about market rents.

I agree that it should show more than 10. 10 is not a suitably large sample and an effect of that is that it might in some cases exclude higher rents. However I have to ask if you have any evidence that it "skews the data to show artificially low rents", or that is just speculation? When I say evidence, either empirical evidence from real searches you have done, or from analysis of the algorithm (which does not contain any criteria regarding rent amount). Excluding top 1% is one factor, I agree, although the stated goal in this case is to elimate incorrectly recorded data, e.g. mistakenly added extra zero, which I think is fair enough.

As I wrote earlier, what I hope to do is be able to analyse this objectively, i.e. replicate the "top 10" and then see what appears in the next 20, next 40 etc.... this would show whether or not higher rents are being systematically excluded from the shown 10 matches because of the arbitrary quantity of 10.

Don't get me wrong, I am not a big fan of the RTB But I am keen to analyse this objectively.

 

[stormy]

However I have to ask if you have any evidence that it "skews the data to show artificially low rents",

It’s more about statistical probability.

Let's assume that in any given 3-month period, only 10% of all RTB registrations are brand-new, open-market lettings. The other 90% are just landlords re-registering existing, capped tenancies. I would think this is a very generous assumption as I don't think 10% of tenants turn over anything like as often as three months.

When the system shows 10 results:
Mathematically, you can only expect 1 out of those 10 properties to be a true market rent. Worse, probability dictates that about 35%* of the time, you will get zero new tenancies in a sample that small. Even if you do get lucky and catch 1 market rent, it is instantly drowned out by the 9 capped rents. The "average" you are forced to look at is clearly below market (as 90% are capped). For the three you need to put on the review/rent setting form, 35% of the time these will all be capped tenancies (i.e. not one of the three will be market rate).

When the system shows 100 results:
You are now capturing a large enough net that you can statistically expect 10 of those properties to be representative, recent market rents. The chance of getting zero market rents in a sample of 100 is virtually impossible (less than 0.01%). The chance of getting at least three market rents is 99.81%**


By restricting the search to 10, the system practically ensures you won't get enough recent data to form a fair benchmark. Increasing it to 100 flips the maths entirely in favor of transparency. Ideally everything should be published.


*35% comes from binomial probability. In the example, we established the premise that only 10% of the RTB database entries are true new market rents, while the remaining 90% are existing/capped renewals. If you pull a random sample of 10 properties, the odds of getting no market rate tenancies tenancies is the probability of pulling a capped rent to the power of 10 = 0.90^10 = 34.867%

**Binomial probability again, I can paste in formula if useful.

 

[patfert1]

It’s more about statistical probability.

Ok, I get it. That does indeed suggest that the key "fault" with the RTB rent register lookup implementation is with the small displayed result set of 10, not with the match score algorithm. The document from the RTB does not mention the display of "10" matches at all.

It is interesting to consider that a lot of thought obviously went into the design of the matching score algorithm, so what was the rationale behind deciding to show just 10 comparables?

 

[stormy]

what was the rationale behind deciding to show just 10 comparables?

The million-dollar question! If they put so much effort into the matching algorithm, why bottleneck the output? Maybe it's an attempt at data privacy - I don't think that's really an issue though as no eircodes published.

Whether this is by design or not, it fits a wider pattern we see with the RTB’s implementation of legislation. For example, the decision to measure annual inflation over 10 months, rounding down to whole months for the 2% cap or including cost rentals in the PRS.

Practically, the legislation is consistently implemented in a way that disadvantages accommodation providers.

 

[Greenbook]

Whether this is by design or not, it fits a wider pattern we see with the RTB’s implementation of legislation. For example, the decision to measure annual inflation over 10 months, rounding down to whole months for the 2% cap or including cost rentals in the PRS.

The rent register is simply part of the wider pattern here that @stormy refers to.

The PRS is not a neutral regulator. It is actively skewing data to support a narrative (landlords aren't leaving) or support a government policy (keep rents below market level).

I think that once 6 months pass from March, the RTB will start 'investigating' landlords who increased rents above the samples.

 

[Gerard0]

Some very interesting comments here on the new Rent Register! I am giving a free talk on Zoom next Saturday at 4pm on the recent changes to landlord and tenant law. Details on Eventbrite Landlord and Tenant Law in Ireland.

 

[Bronte]

Having googled that out of curiosity but cannot find it?

 

[Bronte]

Cork - work in progress

https://www.daft.ie/property-for-rent/cork-city?numBeds_from=1&numBeds_to=1

Big increase in the number of one beds compared to a month ago. 25 available. Discount the studio and housing agency type. Lowest rent is €1250, next one is €1450, 3 at €1700, 4 at €1900 and 5 at €2000, highest is €2200 ! They are all more or less city center except for the ones in Mahon/Rochastown/Douglas & Blackrock.

Let's pop in eircode for one of those - result is Cork City South Central

I'm a bit confused as what it means by South Gate A? Anyone know? Is that a place? 6 are around the 1K mark. The other five are close to 2K. Have to assume the 6 low ones are caught by the rent caps.

 

[galway]

aving googled that out of curiosity but cannot find it?

I think Someone already has linked to it on this site,

 

[Bronte]

Cork other side of the river: Gets us Cork City North East

Is St. Patricks part of the electoral ward? No idea. Compared to the ones in the South of the River the prices are lower. But in actual distance some of the ones on Daft are very close to each other, albeit on different side of the river. 2 are very low rents. The rest are between 1200 and 1600. On Daft DNG have two for rent currently at 1850 + 2000 which match nothing on the RTB.

Learnt something new today as well. On the RTB website you can change the actual electoral area, so gave that a go, found Mahon, rents are lower out there. The Daft one is priced at 1700, none on the RTB, most are well below that, the closest is coming in at 1500.

Now that's pretty handy. Coz one can play with that to justify market rents or any rent actually. My conclusions, the whole thing is nonsense. You cannot actually compare like with like. And anyway, market rent is clearly what the actual new rents landlords are achieving.

 

[Bronte]

Some very interesting comments here on the new Rent Register! I am giving a free talk on Zoom next Saturday at 4pm on the recent changes to landlord and tenant law. Details on Eventbrite Landlord and Tenant Law in Ireland.

Here we go:

Landlord and Tenant Law in Ireland

Join our online session to get the lowdown on Landlord and Tenant Law—know your rights and avoid hassle!

www.eventbrite.ie

Might be of interest to see if we landlords could learn something, or to find out if the experts know more than us !