Having identified the risk, the question of its frequency or magnitude would be very much relevant in insurance.
Consider a factory by the bank of a river causing regular floods and consider another factory near the same river but situated uphill.
Is the risk of flood damage the same for both the factories?
Simple common sense would dictate that the risk of the flood would be more with regard to the first factory (by the bank of the river) as opposed to the second factory (uphill).
To take yet another example consider a house in a comfortable residential area near to a fire brigade office and another house in a very crowdy locality surrounded by lanes and alley bounds and far from any fire brigade office.
Certainly, the possibility of a fire loss would be far higher in the second house as opposed to the first house.
What we are indeed suggesting here is this that in the study of risk we are not simply to contend with the uncertainty as to causation of an event, we should also know the behavioral pattern or risk frequency and its severity as well.
Extend the example of the house by another hypothesis which gives a value to the houses.
The first house in the posh area values $1 million whilst the second house in the crowdy area values $100K.
Now our imagination is a bit changed because we shall have to bring the severity of loss into our scenario.
Because it is the magnitude or cost of a loss also which is of concern to insurers.
Frequency & Severity
As has been indicated in the extended example above, an insurer and risk bearer no doubt we are interested in loss (event) frequency, but at the same time, we are also interested in the severity (cost) of loss.
This is so because ultimately we shall have to pay a loss and our premium generation should be such that would enable us to pay all such claims insured.
Therefore, a correlation is to be established between frequency and severity.
Is it that the more frequent the events are the more is the cost or severity?
This necessarily follows that a distinction is to be drawn between these two.
If we now go through the extended example again can we possibly visualize that although the possibility (frequency) of fire in the house situated at the crowdy fire-prone locality is higher as opposed to the house situated at posh area but the severity of loss, should there be a fire engulfing the house of the posh area, will be much more in comparison to the house of the crowdy area simply because of the higher value involved?
Having said these, when we go for measuring a risk which is necessarily required from the viewpoint of both insurer and the insured we start realizing that a distinction between frequency and severity of risk assumes importance.
This helps insured to decide whether to go for insurance or not.
Similarly, it helps insurer to decide as to what premium would be reason enough to cover loss payment and other incidental expenses, such as, administrative cost, dividend etc.
Let us recall our previous understanding of uncertainty and lack of knowledge about future causation of an event.
The more and more an event occurs our knowledge about future causation of the same event increases and our uncertainty gradually diminishes giving way to certainty.
When uncertainty turns into certainty our prediction about future becomes stronger and stronger and our forecast for future becomes more and more accurate.
This is what an insurer’s objective is and when this point is struck we sit on the driving seat and take the control of forecasting future events as masters thereof.
Going back to the issue of frequency and severity, if a person finds from experience that in his trade or profession the frequency as to the causation of an event is quite high with low cost or severity he might consider retaining the risk of loss on his own shoulder.
On the other hand, if it is found that the frequency as to the causation of an event is rather substantially low with high severity and cost he may transfer the risk to insurers.
Clandestine thefts in private dwelling houses may be one example of high-frequency losses with low cost or severity. Shipping risks, Aviation risks, Petrochemical risks etc.
Maybe examples of low-frequency losses with commendable severity and costs involved.
Following diagrams demonstrates this:
Here the verticle axis represents the frequency of loss event and the horizontal axis represents the severity (cost) of loss.
In private dwelling houses, the incidence of theft is quite high, but the losses are all small clandestine thefts.
What is demonstrated here is this that as the number of incidence or frequency goes up the severity comes down and as the frequency comes down the severity increases.
This position is also supported by a well-known study referred to as Heinrich Triangle.
This was done with regard to industrial injury cases which revealed that the number of major bodily injuries to workmen emanating from industrial accidents is much less as opposed to minor bodily injuries or no injuries at all.
The study was made of workers employed in various industries. The object was to find out the number of bodily injuries arising out of industrial accidents and their severity.
The study revealed that for each major injury there were relatively 30 minor injuries and in 300 incidents there was no injury at all:
This is the normal behavioral pattern of most of the risks.
However, a typical scenario may emerge in rare cases where with the increase in frequency the severity also increases as demonstrated in the following diagram:
Here as the frequency becomes higher and higher the severity also goes higher and higher.
These are normally very high valued risks such as Petro-chemical, Aeroplanes, and Ships etc.
To complete the study of the meaning of risk an understanding of peril and hazard is important.