Seeking input: Return on Equity of Insurance Companies

[David]

The dev team has implemented Return on Equity (ROE) display for banks (see attached)

We would like to see your inputs if you want the same to be implemented for Insurance Companies a well.

One characteristic of insurance companies is that the return from stock investments can fluctuate greatly. It is not uncommon to have returns that exceed 100% or -100% of its equity. For example, in a rather extreme case, if the equity is $10M, and the investment return from stocks is -$200M, then the return on equity would be -2000%. It will disrupt the ROE graph's scale. So currently we cap ROE to a range from -100% to 100%.

Question: Do you have a better suggestion than the current method of capping ROE to a range from -100% to 100%?

---------------------------------------------------------------------------------

On the bank UI, it now also display a 30-year average ROE, which adds up ROE numbers from the past 30 years and divide the total by 30. (or fewer than 30 years, if the bank has been in operation for a shorter period of time.)

But when it comes to insurance companies, the big swings in ROE numbers and the method of capping it to a range from -100% to 100% will render the 30-year average ROE inaccurate. Because of this, the game will not display 30-year average ROE for insurance companies.

Questions:
1) Do you think it is important to display 30-year average ROE for insurance companies?
2) If so, do you have any suggestion for calculating an accurate 30-year average ROE for insurance companies, given the above limitations?

[avengerbg]

the graph can be from -100 to +100 range, while the actual calulations can be without cap, and in case the average goes out of -100/+100 range be capped too, but only for displaying(graph) proposes, game should keep raw and accurate numbers somewhere.

[David]

Example:

Year 1 ROE: -2000% R= -20M E=1 million
Year 2 ROE: -200% R= -2M E= 1 million
Year 3 ROE: 10% R = 1 M E = 10 million

Average ROE: -660%, which does not accurately present the ROE. The low ROE distorts the average.

The culprit: A very small equity value can easily lead to wildly fluctuated ROE numbers.

[avengerbg]

https://en.wikipedia.org/wiki/Normaliza ... tatistics)

May be this can be of some help, used something like in the past for a university project in C++, for statistical data of visitors. Data can be represented in different ways and still be accurate.

https://www.investopedia.com/terms/n/no ... rnings.asp

Here is a bit more info on normalizing ROE, i dont have great if/any knowlidge in economy, but may be you can use this too.

[avengerbg]

here is something from last link that can be very usefull for you:

Understanding Normalized Earnings
Normalized earnings represent a company's earnings that omit the effects of nonrecurring charges or gains. To better present a company's core business, the one-off effects of these profits or losses are removed as they can muddy the picture. Additionally, normalized earnings can be used to present a firm's earnings while taking into account seasonal or cyclical sales cycles.


In short, normalized earnings are the most accurate assessment of a company's true financial health and performance. Many companies incur one-off expenses, such as large lawyer fees, or earn one-off gains, such as the sale of old equipment. In both of these cases, even though the costs and revenues are realized and affect the company's short-term cash flow, they are not indications of the company's long-term performance. To analyze the firm properly, these effects have to be removed.