residual risk
As with risk management in general, handling risk is not a onetime process. Instead, security must be continually maintained and reaffirmed. In fact, repeating the risk assessment and risk response processes is a necessary function to assess the completeness and effectiveness of the security program over time. Additionally, it helps locate deficiencies and areas where change has occurred. Because security changes over time, reassessing on a periodic basis is essential to maintaining reasonable security.
Control risk is the risk that is introduced by the introduction of the countermeasure to an environment. Most safeguards, security controls, and countermeasures are themselves some sort of technology. No technology is perfect and no security is perfect, so some vulnerability exists in regard to the control itself. Although a control may reduce the risk of a threat to an asset, it may also introduce a new risk of a threat that can compromise the control itself. Thus, risk assessment and response must be an iterative operation that looks back on itself to make continuous improvements.
Cost vs. Benefit of Security Controls
Often additional calculations are involved in risk response when a qualitative risk assessment is performed. These relate to the mathematical evaluation of the cost/benefit of a safeguard. For each identified risk in criticality priority order, safeguards are considered in regard to their potential loss reduction and benefit potential. For each asset-threat pairing (i.e., identified risk), an inventory of potential and available safeguards must be made. This may include investigating the marketplace, consulting with experts, and reviewing security frameworks, regulations, and guidelines. Once a list of safeguards is obtained or produced for each risk, those safeguards should be evaluated as to their benefit and their cost relative to the asset-threat pair. This is the cost/benefit evaluation of safeguards.
Safeguards, security controls, and countermeasures will primarily reduce risk through a reduction in the potential rate of compromise (i.e., ARO). However, some safeguards will also reduce the amount or severity of damage (i.e., EF). For those safeguards that only reduce the ARO, the amount of loss of a single realized event (i.e., SLE) is the same with or without the safeguard. But, for those safeguards that also reduce the EF, any single realized event will cause less damage than if the safeguard was not present. Either way, a reduction of the ARO and potentially a reduction of the EF will result in a smaller ALE with the safeguard than without. Thus, this potential ALE with the safeguard should be calculated (ALE = AV * EF * ARO). We can then consider the original asset-threat pair risk ALE as ALE1 (or ALE pre-safeguard) and the safeguard-specific ALE as ALE2 (or ALE post-safeguard). An ALE2 should be calculated for each potential safeguard for each asset-threat pair. The best of all possible safeguards would reduce the ARO to 0, although this is extremely unlikely.
Any safeguard that is selected to be deployed will cost the organization something. It might not be purchase cost; it could be costs in terms of productivity loss, retraining, changes in business processes, or other opportunity costs. An estimation of the yearly costs for the safeguard to be present in the organization is needed. This estimation can be called the annual cost of the safeguard (ACS). Several common factors affect ACS:
Cost of purchase, development, and licensing
Cost of implementation and customization
Cost of annual operation, maintenance, administration, and so on
Cost of annual repairs and upgrades
Productivity improvement or loss
Changes to environment
Cost of testing and evaluation
The value of the asset to be protected determines the maximum expenditures for protection mechanisms. Security should be cost-effective, and thus it is not prudent to spend more (in terms of cash or resources) protecting an asset than its value to the organization. If the cost of the countermeasure is greater than the value of the asset (i.e., the cost of the risk), that safeguard should not be considered a reasonable option. Also, if the ACS is greater than the ALE1 (i.e., the potential annual loss of an asset due to a threat), then the safeguard is not a cost-effective solution. If no safeguard options are cost-effective, then accepting the risk may be the only remaining option.
Once you know the potential annual cost of a safeguard, you can then evaluate the benefit of that safeguard if applied to an infrastructure. The final computation in this process is the cost/benefit calculation, or cost/benefit analysis. This calculation is used to determine whether a safeguard actually improves security without costing too much. To determine whether the safeguard is financially equitable, use the following formula:
[ALE pre-safeguard – ALE post-safeguard] – annual cost of safeguard (ACS) = value of the safeguard to the company
If the result is negative, the safeguard is not a financially responsible choice. If the result is positive, then that value is the annual savings your organization may reap by deploying the safeguard because the rate of occurrence is not a guarantee of occurrence. If multiple safeguards seem to have a positive cost/benefit result, then the safeguard with the largest benefit is the most cost-effective option.
The annual savings or loss from a safeguard should not be the only consideration when evaluating safeguards. You should also consider the issues of legal responsibility and prudent due care/due diligence. In some cases, it makes more sense to lose money in the deployment of a safeguard than to risk legal liability in the event of an asset disclosure or loss.
In review, to perform the cost/benefit analysis of a safeguard, you must calculate the following three elements:
The pre-safeguard ALE for an asset-threat pairing
The potential post-safeguard ALE for an asset-threat pairing
The ACS (annual cost of the safeguard)
With those elements, you can finally obtain a value for the cost/benefit formula for this specific safeguard against a specific risk against a specific asset:
(pre-safeguard ALE – post-safeguard ALE) – ACS
or, even more simply:
(ALE1 – ALE2) – ACS
The countermeasure with the greatest resulting value from this cost/benefit formula makes the most economic sense to deploy against the specific asset-threat pairing.
It is important to realize that with all the calculations used in the quantitative risk assessment process (Table 2.2), the end values are used for prioritization and selection. The values themselves do not truly reflect real-world loss or costs due to security breaches. This should be obvious because of the level of guesswork, statistical analysis, and probability predictions required in the process.
Once you have calculated a cost/benefit for each safeguard for each asset-threat pair, you must then sort these values. In most cases, the cost/benefit with the highest value is the best safeguard to implement for that specific risk against a specific asset. But as with all things in the real world, this is only one part of the decision-making process. Although very important and often the primary guiding factor, it is not the sole element of data. Other items include actual cost, security budget, compatibility with existing systems, skill/knowledge base of IT staff, and availability of product as well as political issues, partnerships,
