Please activate JavaScript!
Please install Adobe Flash Player, click here for download

Allianz Annual Report 2012

Annual Report 2012    Allianz Group Our internal risk capital model covers: −− All of our major insurance operations; −− Our assets (including bonds, mortgages, investment funds, loans, equities and real estate) and liabilities (in- cluding the cash flow profile of all technical reserves as well as deposits and issued securities); −− For the Life/Health segment, the interaction between as- sets and liabilities is driven by local management deci- sions – such as investment strategies and policyholder participation rules. Traditional Life/Health products sold in Western Europe generally feature policyholder participation in the profits (or losses), subject to manage­ ment discretion and typically floored at a minimum guaranteed crediting rate. The majority of our Life/ Health contracts in Western Europe comprise a signifi- cantlevelofpolicyholderparticipation,therebyreducing major sources of risk, including market, credit, under- writing and cost risks;1 −− Almost all of our derivatives, such as options, swaps and futures, in particular if they are part of the operating en- tity’s regular business model (e.g. at ­Allianz Life Insur- ance Company of North America) or if they have a signif­ icant impact on the resulting internal risk capital (e.g. in the Life/Health segment, if material obligations to policy­ holders are hedged through financial derivatives). In general, embedded derivatives contained in a host con- tract are also included.2 For our Asset Management segment we assign internal risk capital requirements based on the sectoral regulatory cap- ital requirements in accordance with the current approach in Solvency II. We allocate these requirements to the risk categories of our internal risk capital model, thereby allow- ing us to consistently aggregate internal risk capital for all segments at the Group level. Approximately 99.8 % of the investments managed by the Asset Management operating entities are held for the benefit of third parties or ­Allianz Group insurance entities and, therefore, do not result in a significant additional market and credit risk for the seg- ment. However, the assessment of market risk and credit 1 For further information about participating life business, please refer to note 20 to the con- solidated financial statements. 2 For further information about additional risk disclosure regarding derivative financial instru- ments, please refer to note 43 to the consolidated financial statements. risk on the account of third parties is an integral part of the risk management process of our local operating entities. The capital requirements of smaller insurance operating entities, that have an immaterial impact on the Group’s risk profile, are either based on local regulatory require- ments or on a risk factor based model. We allocate these to the risk categories of our internal risk capital model. This allows us to consistently aggregate internal risk capital for all segments at the Group level. Internal risk capital related to our European banking op- erations is allocated to the Corporate and Other segment, based on the approach applied by banks under the Basel II standards. It represents an insignificant amount of approx­ imately 1.6 % of total pre-diversified internal risk capital. Therefore, risk management with respect to banking opera- tions is not further discussed. Limitations Our internal risk capital model expresses the potential “worst case” amount in economic value that we might lose at a certain confidence level. However, there is a statistically low probability of 0.5 % that actual losses could exceed this threshold at Group level in the course of one year. We use model and scenario parameters derived from his- torical data, where available, to characterize future possi- ble risk events. If future market conditions differ substan- tially from the past, for example in an unprecedented crisis, our VaR approach may be too conservative or too liberal in ways that are too difficult to predict. In order to mitigate reliance on historical data we complement our VaR analy- sis with stress testing. Our ability to back-test the model’s accuracy is limited because of the high confidence level of 99.5 %, the one-year holding period as well as only limited data for some insurance risk events – such as natural ca- tastrophes – being available. Furthermore, as historical data is used where possible to calibrate the model, it can- not be used for validation. Instead, we validate the model and para­meters through sensitivity analyses, independent internal peer reviews and validation and, where appropri- ate, external reviews by independent consulting firms fo- cusing on methods for selecting parameters and control processes. Overall, we believe that our validation efforts are effective and that our model adequately assesses the risks to which we are exposed. 192