Fund Flows reflect the net of all cash inflows and outflows from a mutual fund or ETF. This means that fund flows do not consider the change of performance from the fund or the assets from the fund. However, a positive fund flow can be interpreted as excess cash for fund managers to invest, which would potentially increase the purchase of assets from the fund.

Fund Flows=fund cash inflows – fund cash outflows

Example: A UK-domiciled investor wants an exposure to emerging markets sovereign debt but wants to diversify by investing in different EM sovereign debt. After some research, he decides to invest £10,000 in a UK-domiciled mutual fund focused on emerging sovereign debt. As a result, the £10,000 would become a fund cash inflow and, if everything stays the same, would reflect a positive fund flow for the mutual fund.

Flow of funds accounts are a system of interrelated balance sheets for a nation, calculated periodically. It is an extensive and reliable source to understand the financial status from specific sectors of the economy such as the government, households, businesses, and financial institutions. For example, we can quantify the level of liquidity risk of the pension fund sector by analyzing its assets and liabilities data provided by the flow of funds. Normally, the government of a particular country publishes the flow of funds quarterly. In the report, there are two types of balance sheets:

• The aggregate assets and liabilities for financial and nonfinancial sectors.
• What sectors issue and hold financial assets (instruments) of a given type.

Cross-border (portfolio) flows are cross-border transactions in financial assets: flows of money and assets across borders. If a UK investor buys bonds or equities from a US holder, that’s a portfolio flows.

Portfolio assets are things like bonds and equities. Economists use the phrase portfolio flows specifically to mean the cross-border flows, and these they record in the balance of payments.

Fund allocations: the asset allocations of mutual funds and ETFs at a point in time are a form of positioning data. If the weighted average allocation of a large group of foreign-domiciled asset management companies’ dedicated emerging market debt funds to South African government debt is below the weight of South Africa in the benchmark index those funds follow, then we say directly that that group is underweight. It might be that, if we see large bond portfolio outflows from South Africa, that that group is reducing allocations to South Africa, but there might be another explanation, for example selling by funds not dedicated to emerging market debt. Positioning data speaks more directly than flow data here.

We calculate average allocations as the average of the sample of available data.

For example, if we find allocations to the Mexican Peso on 35 fund factsheets, then we report the average allocation as the mean of that sample.

The allocations themselves are reported as absolute percentages (for example 9.9% or 8.4%). Average allocations can be calculated for allocations to countries, currencies, duration, yield, cash and credit ratings.

Different definitions of the variables are used to calculate the averages (e.g. Average yield includes YTM, average yields, etc..). Credit ratings average allocations do not sum to 100% as funds often report their allocations in bands, whereas we calculate averages across specific credit ratings.

Since many funds only report their top 10 country allocations on their factsheets, this produces allocation data which has an upward selection bias as we are more likely to include a country’s allocation if it has a higher allocation. This is more pronounced for countries that have a lower weight in the benchmark.

We also calculate average allocations that assume that funds not reporting their allocation for a country have a zero allocation or the same allocation as the benchmark. The closer these values are together, the more confidence we have in the allocation.

Market structure: the composition of the holders of portfolio securities at a point in time are a form of positioning data. Different holders have different motivations and different constraints and these give rise to different behaviours. You can arrive at positioning by knowing about the different investors in a market (market structure) and how they are positioned (e.g. through their asset allocations).