Chris Fawcett Response Document To Remote gambling: financial vulnerability and financial risk

1. Background

My name is Chris Fawcett. I am currently a data scientist and machine learning coder. I have been an avid horse racing fan and gambler for the last 30 years. The majority of my gambling was conducted on the Betfair betting exchange between 2000 and 2015.

In late 2021 I embarked on an experiment gambling using traditional online bookmakers licenced by the GB Gambling Commission. This lasted until early 2023 by which time I had deposited £24,160 into 18 accounts. 1 account remained fully usable. 1 account was closed due to failure to provide bank statements for an affordability check. 1 account was limited to £1,000 a month net loss. The remaining accounts had their stakes limited to pennies on sports bets so were unusable apart from for casino and slots. 13 of the 18 accounts made a profit.

I hope my lived experience from using online betting operators can provide some insights for policy makers in relation to the financial vulnerability and financial risk checks.

2. The State of the GB online market

The Gambling Commission published industry statistics in November 2022 for the year to March 2022. Gross Gambling Yield (GGY) was £2.4bn for online betting and £3.9bn for online casinos. Online slots made up 77% of online casino GGY and was over £3bn for the first time.

The market is dominated by Flutter (Betfair, Paddy Power and Skybet), Entain (Ladbrokes and Coral) and 888 Holdings (William Hill and 888sport). Other big brands include Unibet, Betway, Betvictor and Betfred. These brands all have online casinos and betting. They dominate the online betting market.

The online casino and slots market is more spread out among brands in GB. There appear to be in excess of 100 online slot providers licensed in GB. Furthermore, googling ‘non gamstop slots’ finds 100s more online slot websites with many providing exactly the same games as there are in GB.

3. Financial Vulnerability Checks

3.1. Criteria for checks

Most of my betting accounts would have triggered the £125 loss in a 30 day period check. 7 of the accounts would also have triggered the £500 in a rolling 365 day check.

These checks come in if a loss of £1.37 a day occurs. This appears to be an extremely low threshold.

3.2. What happens when the check is complete?

It is also unclear what happens to the gambler if they have something on their record. Will they be subject to extra monitoring or will their account just be closed? If their account is closed they could just move to the next operator and start again. Slots players, in particular, have 100s of options to play exactly the same product.

3.3. Comments and Proposal

I think these checks will be ineffective unless they are just used for extra monitoring and take place on every account before a bet is placed.

4. Enhanced Financial Risk Assessments

The example given in the consultation was ‘a case which recently led to compliance activity by the Gambling Commission, a customer lost £36,000 in four days without appropriate financial risk assessment being carried out.’

4.1. Criteria for checks

The example given seems an extreme case. Of the 18 online betting accounts I used 14 would have met the criteria for either (or both) £2,000 lost in a rolling 90 day period or £1000 in a day. This is despite all the accounts combined never having a losing month and the average annual deposit per account being £1342. Sports betting naturally has variance in results which means that using rolling time scales means eventually even a relatively small gambler will meet the criteria on an account by account basis. See Type of Gamblers report for more details.

The criteria works slightly differently for slots players. They are more likely to have less variance and just steadily lose over time. The lack of variance means they are less likely to hit a threshold even if they lose more money than a sports gambler over the long term.

Once the threshold is hit the gambler will either submit to the check or refuse. I was subject to a check being asked for bank statements and pay slips by one operator. I refused because the operator refused to say who would have access to the information or how it would be stored. They terminated the account and offered no information on gambling harm. Another operator engaged and offered me a £1,000 30 day rolling loss limit. This included bets already staked so rendered the account useless to me. I carried on with the other 16 unchecked accounts.

The White Paper states that just 3% of accounts will require these checks. This is calculated from 2020/21 data before the country was hit with inflation. I estimate that just accounting for inflation, 4% of accounts will be checked if the thresholds are not increased. For the case of horse racing I estimate that accounts that contribute between 70% and 90% of online GGY will be checked. A survey by the BHA of 14,000 people found that 26% of respondents have already been asked for checks. I would expect this percentage to increase dramatically if the current thresholds for checks on losses are imposed.

4.2. What happens when the check is complete?

As with the Financial Vulnerability Checks it is unclear what happens even if a gambler does agree to checks. The checks will not be a pass or fail. If a gambler has lost £1,000 in a day or £2,000 over 90 days I assume the idea is they will then be given some kind of deposit or loss limit based on the information obtained by the operator.

Frictionless checks through credit reference agencies are extremely unlikely to be able to provide enough information to accurately assess if a gambler should be given a deposit limit or loss limit above £2,000 every 90 days. In order to get a higher limit more information will be required.

Open banking presents the gambler with equal or greater privacy issues than submitting information directly to the betting operator. It is another person/people and computer system in the loop. Even this information will not be enough for people who own their own businesses, are retired or have other special circumstances. I expect between 50% and 85% of gamblers will refuse to provide this extra information. They will either have their accounts closed or restricted to point well below their wishes.  

The reality is if gamblers have their accounts closed or restricted to a point below what they want to bet they will move on to the next operator. Once they run out of GB operators (would take a long time in the case of a slots player) they could consider ‘non-gamstop’ operators. Everytime they go to a new operator the picture of risk starts again. A gambler could still lose tens of thousands of pounds in a short space of time.

In sports and racing betting there are less operators compared to slots gambling. As my experiment demonstrated, sports betting accounts are also subject to heavy restrictions if they show any signs of winning. There are relatively few licensed operators that allow sharp gamblers. For instance, if a gambler does not want to give intrusive documentation to the Betfair, Matchbook and Betdaq exchanges they might effectively be cut out from the entire market. This seems very unfair when a slots gambler has the option of 100s of operators to move on to.

4.3. Comments and Proposal

I think that if the goal is to stop large losses like the example above a less complicated system should be applied:

1. A net deposit limit for a rolling 90 days would be much easier to keep track of. It would allow gamblers to save money in their betting account before a big event and build up a betting bank specifically for their hobby. It would also encourage gamblers not to move from provider to provider where their risk would have to be assessed again.

2. Slow down the process of registration so gamblers cannot lose tens of thousands of pounds very quickly. I suggest on registering a new account all identification and financial vulnerability checks are carried about BEFORE depositing and betting is allowed. As with the deposit limit system it would allow more conscious gambling.

5. Data Protection for Information Collected For The Checks

SR Code 3.4.5 says:

‘In line with data protection principles, licensees must only use the data obtained pursuant to paragraphs 4 and 5 above to consider the risk to the customer and decide what proportionate action to take and it must not be used for any other purpose.’

This is a noble endeavour but it is impossible to visualise how this will work in reality.

5.1. The Status Quo

The status quo in the industry is that operators regularly ask customers for bank statements, p60s, inheritance certificates, mortgage documents and other private information. In my experience it is never made clear exactly why this information is required except a general comment such as ‘Due to our Gambling Commission licence you must submit…’.

The information is sometimes asked for ‘source of funds’ and ‘kyc’ checks but it is not made clear if this is for affordability checks (financial risk checks), AML checks or to enforce private terms and conditions (i.e. commercial purposes).

Currently some betting operators refuse to say how documents and data will be kept secure, who will have access to them and for what reason. Some operators are licensed by the Gambling Commission but are based outside the UK so it is unclear what powers the ICO will have over them to make sure information is not used for commercial purposes.

Wes Himes from Betting and Gaming told the DCMS committee that ‘Black market sites would try to poach the databases from VIP managers and operators in order to contact those people and steer them towards the site.’ Whether they have been successful or not was not stated but it does bring into question how secure the data is and who has access to it.

5.2. Comments and Proposal

In order to build trust there needs to be proper systems in place to protect customer data from misuse. I suggest that:

1. No information required due to licensing conditions can be used for commercial purposes rather than just information required for financial risk checks.

2. When requesting information betting operators must state if it is for commercial purposes or for licensing purposes.

3. Operators must provide customers with the name of a human point of contact that is responsible for their data and information.

4. Everyone that accesses the data and information should be logged so if a data breach occurs it can be audited.

6. Conclusions

The status quo where all operators are carrying out their own checks requiring all kinds of information from customers is chaotic and unsustainable.

I suggest that if a form of affordability check is required by the government it is done in a much simpler manner than proposed. The original remit was to prevent people losing five figures in a relatively short space of time.

The current proposals will do nothing to stop this. With over 100 GB licensed slots operators and 100s more with the same games overseas potential gambling addicts could just move from operator to operator. If they are blocked by some operators there are plenty more. The same applies to sports and racing gamblers, but to a lesser extent.

I think the worst thing the proposals could do is force operators to close gamblers accounts for failure to provide evidence of having the funds to bet without offering any support. I suggest more engagement with potential addicts by betting operators and a slowing down of gambling, especially with new accounts.

Net deposit limits rather than loss limits will be far easier for industry and gamblers to keep track of. It will allow sports and racing gamblers to bet more consciously and plan how much of their money they want to set aside for gambling. It should also reduce the number accounts that require checks just due to short term variance.

Before any formal check is required consideration should also be given to length of time an account has been open for and if anything has changed in the pattern of use of the account. For instance, if an account has been used for decades and nothing has changed in its pattern of use it should be unnecessary to do a formal check of affordability.

If gamblers are to be required to allow betting operators access to their private financial details there must be cast iron systems in place to protect it. Betting operators that have been fined over £100m by the Gambling Commission for bad behaviour have to earn that trust. The Gambling Commission also has to show it has the enforcement power to prevent the use of data for commercial purposes. Without this trust the majority of gamblers will refuse to comply. They will stop betting or go to non-GB licensed operators.

The variance in sports betting results provide an increased risk for betting operators keeping accounts open under the proposed framework. Sports and racing betting are also likely to require far more checks and administration than slots betting customers. This means we are likely to see a further shift from sports and racing to slots betting in the GB market as a result of these regulations. Less opportunities for sports and racing gambling in the GB market will mean more gamblers seeking out non GB licensed opportunities. Slots players are far less likely to suffer checks, due to lack of variance, and have hundreds of substitutes if banned from one operator.

These unintended consequences of regulation will likely see a shift from betting to more addictive slots and an increase in the black market, whilst providing no practical limits on what gamblers can actually lose. We could see an increase in gambling harm, a decrease in tax revenue and the destruction of British horse racing as a result of these well meaning policies.

Consultation Questions and Answers

67. To what extent do you agree with the proposal that gambling operators be required to conduct light touch financial vulnerability checks based on public data when a certain net loss threshold is reached?

see Section 3.1. in attached document

69. To what extent do you agree with the proposal that gambling operators be required to conduct enhanced financial risk assessments where there are very unusual patterns of loss? The purpose of such an assessment would be to act on the indicator of harm of unusual patterns of loss and assess gambling in the context of a customer's financial circumstances.

see Section 4.1. in attached document

71. To what extent do you agree with the proposed threshold of a financial vulnerability check based on public data (eg bankruptcy) if a customer has a net loss of £125 in a rolling 30 day period?

see Section 3.1. in attached document

72. To what extent do you agree with the proposed threshold of a financial vulnerability check based on public data (eg bankruptcy) if a customer has a net loss of £500 in a rolling 365 days?

see Section 3.1. in attached document

73. To what extent do you agree with the proposed threshold for a financial risk assessment related to binge activity of more than £1,000 in a relevant period of a rolling 24 hours?

see Section 4.1. in attached document

74. To what extent do you agree with the proposed threshold for an enhanced financial risk assessment related to significant losses over time of more than £2,000 in a rolling 90 day period?

see Section 4.1. in attached document

75. To what extent do you agree with the proposal that thresholds for the enhanced financial risk assessment are lower for those aged under 25 to £500 in a rolling 24 hour period and £1,000 in rolling 90 day period?

In the UK people are able to vote, marry, be an MP and be in the military at 18. It seems very patronising to treat adults under 25 as lesser adults just for gambling

76. To what extent do you agree with the proposed definition of net loss for financial risk assessments which is that net loss is the loss of deposited funds with a particular operator ie excluding bonus funds and restaked winnings? In particular, please flag any potential risks and technical difficulties with implementation of this proposed approach.

In my experience many operators do not keep these records up to date. Skybet, for instance, would not show aggregated profits and lose data at weekends until well into the next day. If blanket limits are deemed the best methodology then I think deposit limits rather than convoluted  losses would be far easier to manage and track.

see Section 4.3. in attached document

77. To what extent do you agree with the proposed approach to enable a recent overall net position to be taken into account when a threshold is met? In particular, please flag any potential risks and technical difficulties with implementation of this proposed approach.

As 76. Unnecessarily complicated.

see Section 4.3. in attached document

78. To what extent do you agree with the proposed approach that would set a timeframe whereby recent overall net position could be taken into account for 7 days in relation the binge threshold and 90 days for the losses over time threshold?

These are arbitrary numbers. 7 days for someone betting £2 a spin 100s of spins daily on slots is different to weekend gambler betting on football and racing.

79. To what extent do you agree with the proposed approach that would mean that a bet would only be counted as a loss when it is settled as a loser?

This is another example of the issues with using losses rather than deposits to measure harm. If bets are not counted as a loser when placed then an addict could potentially bet tens of thousands on a Saturday afternoon and lose it all. If the goal is to check the gambler can afford to lose the money then this system would be an abject failure. At the same time if a gambler wishes to bet on something in the future like a team to win the Premier League at the start of season it would be ridiculous to associate this with binge gambling or sustained loss over time.

80. To what extent do you agree that a financial vulnerability check would include publicly available data relating to an active county court judgement (CCJ), high court judgment (HCJ), administration order (AO) or decree, or equivalent?

It is unclear what the operator is supposed to do with this information.

see Section 3. in attached document

81. To what extent do you consider that aggregated data should be included in a financial vulnerability check in relation to postcode?

I think this is deeply discriminatory and should not be used.

82. To what extent do you consider that aggregated data should be included in a financial vulnerability check in relation to a customer’s stated employment status and job title, and cross-referencing to open source data about the average income for that occupation?

I think this is deeply discriminatory and should not be used.

83. To what extent do you agree with the proposed requirements for data that must be included in an enhanced financial risk assessment for credit performance data and income and expenditure data, including current account turnover data?

84. Please give your reasons for your answer. In particular, are there any other types of information that you think it would be valuable to gather at these thresholds to understand potential financial risk?

Income, expenditure and CATO data is very limited. It is unclear how this helps an operator take action to prevent gambling harm. It does not include savings and other assets so does not give a full picture of how much someone can ‘afford’ to bet. It is also only a snapshot of a point in time.

see Section 4. in attached document ‘

85. In limited circumstances, it may be necessary to obtain information directly from the customer to understand financial risk. In these circumstances, we have proposed that the information must enable assessment of income and expenditure. Should the Commission set out further minimum requirements to ensure the data provided is meaningful but minimised? If so what should these requirements be?

To get a complete picture of someone’s affordability information on savings, relationship status, assets (such as home ownership), dependants such as children and also if they intend to have children in the near future. It is simply too intrusive that most people will not be willing provide it just to have a bet.

86. To what extent do you agree with a 12 month time-frame for the validity of the financial vulnerability check?

Circumstances can change dramatically in a day let alone a year. A sudden change such as a new relationship, break down of a relationship, losing a job, getting diagnosed with an illness can change someone’s affordability situation overnight. A change for the worse could also make someone more vulnerable and more likely to gamble as an ‘escape’.

87. To what extent do you agree with a 6 month time-frame for the validity of the financial risk assessment?

See 86.

88. To what extent do you agree that it is proportionate that deposits and gambling may continue while a financial vulnerability check is taking place?

If this check is deemed necessary at such a low level that it should take place BEFORE any betting takes place on all accounts.

89. To what extent do you agree that it is proportionate that gambling may continue while a financial risk assessment is taking place, but that further deposits would be paused?

It depends on the circumstances. If there are markers of harm such as multiple daily deposits and signs of chasing loses then contact should be attempted and stakes should be restricted. For an established customer following their regular betting patterns then of course gambling should continue. They should not be facing a check in the first place.

90. We have not proposed any set requirements for how quickly a financial vulnerability check must be completed. Do you have any comments on whether such requirements would be necessary?

Immediate. If they are deemed necessary they should take place before deposits are made and any gambling is allowed.

see Section 3.3 in attached document

91. Please provide any views you may have on the best way for gambling operators to inform customers about the potential collection of their financial data for these purposes.

In my lived experience, checks can take over 2 months to be completed.

It is more important that users are told exactly why checks are taking place. Currently betting operators say checks are required for ‘KYC’ or ‘Source of funds’ but do not state if they are for affordability, AML or private terms and conditions.

see Section 5. in attached document

92. What feedback do you have on the requirement on operators for manual review of the assessment data, together with all of the other information they hold on the customer to make a proportionate decision on any action to be taken?

see Section 4.2 in attached document

93. Does paragraph 7 of the proposed requirement 3.4.5 (enhanced financial risk assessment) which confirms that operators can only use the information collected for the purpose of the assessment, provide sufficient clarity that the information must not be used for any other purpose?

see Section 5. in attached document

94. What factors should be considered in relation to implementation timeline and piloting?

The current situation is a total shambles with operators demanding intrusive information without stating if it is for commercial purposes, AML checks or affordability. The Gambling Commission needs to get a grip on this and make sure all data is secure before embarking on a pilot for the new checks.

see Section 5.2 in attached document

95. Do you consider that there should be any specific appropriate record-keeping requirements?

see Section 5.2 in attached document

96. Do you have any evidence or information which might assist the Commission in considering any equalities impacts, within the meaning of section 149 of the Equality Act 2010, in the context of the proposals set out in this section of the consultation relating to light touch financial vulnerability checks and enhanced financial risk assessments?

No

97. If you have relevant information, please provide an estimate of the direct costs associated with implementing the light touch financial vulnerability check.

No

98. If you have relevant information, please provide an estimate of the direct costs associated with implementing the enhanced financial risk assessment.

No

99. Do you have any comments on the assumptions underpinning the impact assessment set out in Annex A of the white paper?

Yes. The calculations used by the DCMS to estimate the cost to horse racing were flawed and likely underestimate the costs to racing by at least 10 times.

see the following documents:

https://gamblingreform.co.uk/dcms-financial-risk-checks-impact-racing-assumptions

https://gamblingreform.co.uk/static/reports/risk-checks-racing-impact-assessment.pdf