Analyzing ETF Flow Trends for Trading & Investment Analysis

Published May 21, 2025 – Aniket By Aniket Ullal, SVP and Head, ETF Research & Analytics

Key Takeaways

ETF flows measure the value of investment dollars that investors have either added or removed from an exchange traded fund.
The net flow calculation reflects the changes in the shares outstanding of the ETF based on investor demand.
ETF flows can be a useful input into assessing investor risk on / risk off sentiment. Flows can also be analyzed through the lenses of sector, strategy and theme.
However, incorporating nuances such as ‘create to short’ and the impact of model portfolios is important when using flows data in trading analysis.

The Definition of Exchange Traded Fund (ETF) Flows

ETF flows measure the value of investment dollars that investors have either added or withdrawn from an exchange traded fund (ETF). It is important to note that flow is not the only factor that increases or reduces the assets in an ETF. ETF assets also change due to changes in the market value of the existing securities held in the ETF.

In other words, the value of assets in an ETF can change for two reasons – net flows (i.e. new investor money coming in or out of the fund) and market movement (i.e. the change in the market value of the underlying securities already in the fund).

Let us explore this with a very simple example: Let’s assume an ETF has $100M in assets at the end of Day 0. On Day 1, let’s assume net inflows of $9.5M, i.e. investors put in $9.5M of new money into the ETF. Let’s also assume that the value of the securities already held in the ETF goes down by $5M due to a market decline.

So, there are two offsetting changes happening to that ETF on the same day – the value of assets is going up due to a new inflow, but assets are also declining due to market movements. Therefore, on Day 1:

Table 1: Components of asset change (net flows and market movement).

Therefore, the net assets at the end of Day 1 will be $100M + $4.5M = $104.5M.

Figure 1 illustrates a real-life example of how net assets for an ETF are impacted by both flow and market movement. The red arrow points to the asset change for the SPDR Gold Trust (GLD) on April 22, 2025.

Figure 1: Asset Change in GLD on Apr 22, 2025, due to Flows and Market Movement

Source: CFRA’s FUNDynamix ETF analytics portal

The dark blue shows the change due to market movement, i.e. the gold held in GLD appreciated in value by $3.94 billion on that day. However, net flows due to investor redemptions (shows in the light blue) were -$1.27 billion. Therefore, the assets of the ETF on that day went up by $2.67 billion (shown by the dot) i.e. $3.94 in market appreciation – $1.27 of net flows.

In these examples, market movement and net flows partially offset each other, but that doesn’t always have to be the case. On a given day, both market movement and flows could be positive (or negative).

How ETF Flows Are Calculated

On a given trading day for an ETF, net flows are calculated as follows:

Net flows on a trading day = Change in shares outstanding on that trading day * NAV at the end of that trading day.

The net flow calculation reflects the changes in the shares outstanding of the ETF based on investor demand. If new money is invested, the shares outstanding will increase. If money is redeemed, the shares outstanding will be reduced. This calculation is adjusted for corporate actions such as splits or reverse splits.

Let us look at this calculation in a little more detail, using our earlier hypothetical example of the $100M ETF. Let us assume that at the end of Day 0, the ETF had 5 million shares outstanding at an NAV of $20. i.e. assets of $100M.

On Day 1, 500,000 new shares were created (i.e. there were inflows) but the market declined i.e. NAV fell from $20 per share to $19. This is summarized in Table 1 below.

Table 2: Hypothetical example of an ETF with inflows and market decline.

Figure 2: Sample ETF Watchlist to Track Investor Risk On / Off Sentiment

To calculate the flows, we use our flows formula i.e. Change in change shares outstanding on that trading day * NAV at the end of that trading day.

However, the underlying value of the securities held declined i.e.

The net asset change is a sum of these two effects, i.e.

Separating out net flows from market movement provides useful insight, because it enables analysis on investor demand, which is independent of changes in assets due to price movements.

Flows for ETFs vs Mutual Funds

When analyzing ETF flows, it is useful to distinguish the flow process for ETFs and mutual funds, and how they are different in a few fundamental ways:

In-Kind Creation Redemption: In the ETF structure, shares of the fund are created or redeemed by Authorized Participants (APs), which are designated firms that trade directly with the ETF
issuer. When creating new ETF units, the APs will hand in a basket of constituent shares to the issuer, and in return, the issuer will give the AP a block of ETF shares (called a creation basket). These shares then list on an exchange where individual investors or firms can trade them. Similarly, when redeeming fund units, the AP will hand in a block of ETF shares to the issuer and in return receive an equivalent basket of underlying constituent shares (redemption basket). This exchange of ETF units for underlying securities is called in-kind creation and redemption. In contrast, a mutual fund creation or redemption happens in cash when individual investors buy or sell fund units. The in-kind mechanism is more tax efficient, giving ETFs a tax advantage over mutual funds.
Daily data vs monthly data: ETFs trade intra-day and for most ETFs in the U.S., the underlying basket of securities, NAVs and shares are published publicly by ETF issuers. For most mutual funds, it is more difficult to source daily shares, assets, and holdings, and so flows are typically calculated using monthly data. This means that mutual funds flow data tends to be less granular and timely than ETF flow data.
Institutional vs retail user base: Mutual funds tend to be used more by retail investors, though many funds also have separate institutional share classes. In contrast, ETFs only have one share class in the U.S. that are used by both retail and institutional investors. For this reason, ETF flows reflect both retail and institutional behavior, with the latter generally reacting more quickly to market events.

Gauging Sentiment Using Flows Data

ETF flows can be a useful input into assessing macro level investor sentiment. Figure 1 is a custom screener created in CFRA’s FUNDynamix platform. It has a set of ETFs that are useful proxies for gauging investor risk on / risk off sentiment. Some of the ETFs like the iShares Russell 2000 ETF(IWM) and the Technology Select Sector SPDR (XLK) had year-to-date outflows as of May 16, 2025. Typically, these ETFs tend to have inflows when investors are in a more growth-oriented, risk-on mindset. ETFs like the SPDR Gold MiniShares Trust (GLDM) and the Utilities Select Sector SPDR (XLU) had inflows during the same period, reflecting defensive positioning by investors concerned about tariffs and inflation.

Figure 2: Sample Watchlist to Track Investor Risk On / Off Sentiment

Data from CFRA FUNDynamix ETF platform as of May 16, 2025

Monitoring risk sentiment using ETF flows for different time periods can be a valuable input for analysts and portfolio managers looking to assess current investment sentiment.

Flows by Strategy, Theme & Sector

In addition to assessing investor sentiment, ETF flows can also be viewed through the lenses of sector, strategy and theme. For example, Table 3 shows the flows into sector focused ETFs in the U.S. in calendar year 2023. In that year, investors were starting to get back into risk-on mode after the equity downturn of 2022. In March 2023, Chat GPT-4 was released, accelerating investor flows into Technology ETFs as well as other growth-oriented sectors like Consumer Discretionary and Communication Services. This flow activity effectively highlights the investor interest in the Mag-7 trade, since those 7 stocks were among the biggest holdings in these three sector ETFs.

Table 3: Flows into Industry Sector Focused ETFs in Calendar Year 2023.

Data from CFRA FUNDynamix ETF platform as of May 16, 2025

Table 4 provides another example of flows data reflfecting investor behavior. It shows the ETF flows into bond ETFs in the U.S. by maturity category in 2022. The Federal Reserve started an aggressive rate hiking cycle in March 2022 and in response, investors began moving more money into short duration bond to lock in higher yields and to be in less rate sensitive segments of the yield curve.

Table 4: Flows into Bond ETFs by Maturity Category in Calendar Year 2022.

Data from CFRA FUNDynamix ETF platform as of May 16, 2025

Nuances in Using ETF Flows as a Trading Signal

As we have seen, flow data can provide useful insight into overall risk on / risk off sentiment as well as investor rotation across sectors, themes and strategies. However, it is important to note several nuances when analyzing ETF flows data:

‘Create to short’: As described earlier, flows data is calculated using changes in shares outstanding. However, sometimes shares could also be created for the purpose of shorting an ETF. So flows data should always be examined in conjunction with other variables such as short interest data, to confirm whether the flows genuinely represent investor demand rather than short selling.
Relationship to performance: Professional investors often try to assess whether flows are predictive of future performance. The results are often mixed and sometimes flows can lag performance, when investors try to chase returns. The relationship between flows and performance can vary by asset class and time-period, and it should not be assumed that flows are always predictive of performance.
Lag in retail data: Retail flows into ETFs can often be slower than institutional flows. This is because retail portfolios rebalance less often, and model portfolios, which often drive advisor activity, are usually rebalanced monthly or quarterly. Model portfolios, particularly those from large model providers like Blackrock, can drive significant flows when they rebalance. For example, Figure 3 shows the flows into the iShares MSCI USA Quality Factor ETF (QUAL) over the trailing three years through May 17, 2025. The yellow arrow highlights an inflow of $7 billion that came into the ETF over a two day period due to the addition of the fund to Blackrock’s model portfolio in March 2023.

Figure 3: Flows into QUAL In the Trailing 3 Years

Data from CFRA FUNDynamix ETF platform as of May 17, 2025

Incorporating all these nuances such as ‘create to short’ and the impact of model portfolios is important when using flows data as a signal in trading models.

CFRA’s ETF Flows Data Set

CFRA offers a comprehensive set of data, ratings and research to track the global ETF industry. The ETF data consists of three components – constituent holdings, proprietary classifications, and daily statistics. The latter includes daily flows for individual ETFs, available via both feed and API. Flows data can also be accessed and analyzed using CFRA’s FUNDynamix ETF platform.

Trial access for CFRA’s ETF data and tools can be requested here.

Related Insights & Articles

ETF Comparison, Rankings and Ratings

With ETF assets at $10.4T in 2024, picking the right funds is key. CFRA’s machine-learning ratings offer fast, forward-looking insights on cost, risk & returns to boost portfolio performance.

Read Full Blog

Best Defensive Stocks and ETFs to Protect Your Portfolio

In this recording, CFRA’s Vice President, Equity Research Alexander Yokum, CFA, highlights that the recent unemployment uptick is due to workers joining the labor force rather than layoffs.

Read Full Blog

When Is The Best Time To Buy Bonds?

The U.S. health care industry endured challenging conditions in 2023, including persistently high inflation rates, labor market struggles, and recovery from the Covid-19 pandemic.

Read Full Blog

Asset Allocation During Economic Uncertainty

Strategic Asset Allocation for Economic Uncertainty

Recession-proof your portfolios with asset allocation strategies. Balance equities, bonds, and cash to stay ahead during economic uncertainty.

Read Full Blog