We use modern portfolio theory without adjustments for "beta", reverse optimizations, incorporation of investment manager views, or various security parameter averaging practices. Either individually or in combination, these approaches shift portfolio analysis and optimization into a virtual realm, completely detached from reality.
Bond Allocation – We work exclusively with low- to moderate-duration U.S. dollar-denominated bond ETFs. Their credit rating has to be not lower than one notch below investment grade (low credit risk), and they should not be exposed to non-US dollar currency risk. This makes them relatievely comfortable during periods of market stress, without fearing a significant drop when markets run to safety. The only exception is the high-yield bond ETF, whose components' ratings can be significantly below investment grade. However, due to its specific nature, it is closer to equity than bond ETFs. We view bond asset classes as defensive, rather than speculative, instruments.
Unlike many other portfolio modeling services that exclude alternative investment segments, particularly commodities (due to their poor performance over recent decades), the IskraIndex methodology enables the assessment of their expected returns for comprehensive optimization alongside other asset classes. The core premise of the IskraIndex is that the risk of any asset should be compensated by returns, and sooner or later, the negative trend in the commodities sector will reverse. The IskraIndex swiftly incorporates new asset classes into its optimization framework once corresponding ETFs become available. Notably, cryptocurrencies have emerged as a significant new asset class, with ETFs introduced in 2024, which are now included within the IskraIndex optimization scope.
During certain periods – even in the absence of market stress – our model may allocate 100% of the portfolio to bond ETFs, which may be attributed to the extreme overvaluation of riskier asset classes.
