We are back in a zero Fed Funds rates environment, similar to the 2011-2015 period and the brokerage industry, in general, is hurting.
Figure 1: Fed Funds Rate from fred.stlouisfed.org
While Interactive Brokers Group (NASDAQ:IBKR) only invests in very short-term treasury notes out of prudence – making the Fed Funds the appropriate benchmark – IBKR’s American competitors (Charles Schwab Corporation (NYSE:SCHW), TD Ameritrade Holding (NASDAQ:AMTD)) lend long in various products such as municipal bonds and mortgages. Their benchmark is more dire when compared to ’11-’15:
Figure 2: 10-Year Treasury rate from investing.com
It is only logical that consolidation continues in such a situation to synergize costs. Competitor margins were lower before COVID-19 happened. Less consumer options mean a higher growth percentage of “new customers” for survivors, everything else equal. Indeed, IBKR commented it was seeing an uptick in accounts growth after the fall 2019 mergers of Schwab + AMTD and Morgan Stanley (NYSE:MS) + E*TRADE Financial (NASDAQ:ETFC), pre-COVID-19:
Figure 3: Author’s own graph based on IBKR’s monthly metrics filings. Last account growth peak in the figure buoyed by mergers, note the chart stops before COVID-19 impact in March (previous peak was driven by IBKR’s i-broker Tiger in China)
So, before dabbling in Excel models, let us not forget the difficult-to-quantify but real beneficial effects that zero rates pressure has on industry structure and barriers to success.
Consensus earnings estimates
The following table contains the consensus earnings estimates from Bloomberg as of today:
In a nutshell, 2021 will be lower as:
- the full effect of lower rates will be felt (in 2020, IBKR got the benefit of one quarter of higher rates + delayed effect of full roll-off of higher yielding treasury bills bought in Jan & Feb)
- the trading frenzy which supposedly runs off. Note that while 2020 might ‘feel’ unique, 2009 was also a year of trading frenzy while rates went down: this is a nice counter-balancing effect, helping EBIT to remain constant in 2020 despite a growing employee cost base
- negative operating leverage as IBKR’s opex is always growing on a smaller revenue base
I will show why these numbers to ’22 miss the long-term forest for the one-off interest rate tree.
I model IBKR’s two main gross profit sources – net interest margin (NIM) and commissions. These are net of interest costs (if any) and execution costs to exchanges.
NIM and commissions were roughly 50/50 of gross profit in 2015, but 67/33 in 2019.
I then deduct all (growing) operational expenses (IBKR is always growing employee base) to find underlying operating profit.
An important driver of NIM is, of course, Fed Funds rate as both margin loan rates and treasury bill yields are pegged to this rate (IBKR invests all leftover customer cash balances after loaning some of them as margin loans to other customers in the US treasury bills).
I will assume rates stay at zero in this exercise.
IBKR defines % NIM (or NIM spread) as NIM as a percentage of all interest-earning assets (MUSD). A history from filings:
Figure 4 Author’s compilation of IBKR 10-K filings, section “net interest margin”.
We know 2015 was a full year of 0.0% Fed Funds rate, so without mix effects in client equity and changes in international interest rates, NIM % should stabilize at this 2015 level, give or take, in a zero-interest rate environment. Is it safe to assume everything else equal?
Mix effects & international interest rates
Now, of course, there are mix effects. What are these? Margin loans, cash balances, and both small accounts and intro-brokers account as % of client equity should lead to a higher NIM % spread. In other words, large and boring clients (think financial advisors AUM) are less profitable in terms of spread. Large because smaller clients pay much higher margin rates (see here). And boring because IBKR does not earn one penny of NIM gross margin from a 100% long-only client without any cash and margin debt (except if he allows the lending out of his shares to short-sellers for a 50% cut).
So, what has changed since 2015?
The more profitable client type introducing brokers gained ~10 pp share in the mix to 26% of client equity, taking this share mostly from IBKR’s least profitable client as a percent of client equity. Introducing broker clients really stand out in terms of profitability because of all the factors listed above. IBKR’s other client types such as hedge funds, prop shops, and individuals remained relatively stable (less than 1 pp shift).
However – a rather small (est. 25%, see Q1 2020 conference call) – part of IBKR client cash balances are held in currencies all over the world and the bad news is that funding rates have generally gone down further, which has compressed IBKR’s overall NIM spread as a result:
Figure 5: From FT.com. Major Central Bank funding rates are down from 2015.
This is why IBKR NIM spread was 0.99%, a comparable figure to 2015 as shown in the figure above (a fall of 40% YoY versus Q2 2019):
Figure 6 From IBKR’s 2019 10-K. Net interest margin in Q2 2020 versus 2019.
One last negative and extraordinary effect in Q2 was the fear that led investors to take off margin loans. IBKR invests all customer cash that cannot be lent to other customers to short-term treasuries (i.e. less profitable):
Figure 7: Author’s calculation from IBKR monthly metrics.
Figure 8: Author’s calculation based on IBKR filings. Large but short pullback in client margin loans.
From all the above, it is clear there are a lot of moving parts, but I believe it is reasonable to believe NIM spread will remain at 1.0%, just as it did in the prior US zero rates period. While Q2 2020 was at 0.99% and there is yet a small negative roll-off effect of older treasuries in IBKR’s average ~45-day treasury portfolio, this effect is very small (I estimate 0.03%) and fully offset by the mix shift away from treasuries from Q2 into next years towards the more profitable margin loans. Indeed, Q2 was characterized by fear and abnormally low levels of margin loans. Margin loans just need to revert to their 2019 levels in the mix to fully offset the negative delayed effect from zero rates on the treasury portfolio.
If we would annualize zero rates over 2020 including Q1 2020, IBKR makes ~800MUSD in gross profit from NIM. I will call all my earnings estimates under a zero rates environment “underlying earnings”.
NIM growth over time
As NIM spread remains at 1.0%, NIM gross profit dollars grow with client equity.
Client accounts have grown 16% CAGR since IPO and have not shown slowdown. The low churn (not published but sometimes commented on). Client equity growth results from new accounts, new deposits on existing accounts, and asset returns on those accounts (more volatile). Because of all three, client equity growth since IPO has been 29.5%.
The currently unprecedented new accounts growth will have a one-off boosting effect on client equity which I will model.
Figure 9: Author’s graph based on IBKR monthly metrics filings. IBKR net accounts growth at unprecedented levels.
As the figure below shows, client equity growth is not slowing. Because of the work-from-home environment, net accounts growth is currently at unprecedented levels, having grown 47% YoY already by September 2020. It is true the average rate was boosted by taking the aftermath of the financial crisis as the starting point.
Figure 10: Author’s graph based on IBKR annual filings. Growth in IBKR’s key KPIs.
For the next 10 years, I believe client accounts can grow at 15% and client equity can grow at 23%.
NIM gross margin should, therefore, grow 23% CAGR in a zero-rate environment.
There will be a one-off boost from the stratospheric new accounts won during COVID-19 measures in 2020, as deposits for new accounts come in (typically a good deal of total funding in new accounts is in the second year). Client equity should, therefore, see upward pressure for at least the next 12 months.
I model a linear reversion of accounts growth from the last reported September number to my long-term 15% annualized growth estimate by March 2021 “vaccine”, the end of the work-from-home.
Figure 11: Author’s model. Actual numbers (in dark colours) based on calculations from IBKR filings.
This simple approach allows us to see that total accounts will have grown from March ’20 to March ’21 (exactly one year) by 47% YoY instead of 15%. A one-off boost of 28% (1.47/1.15 – 1) in account and hence client equity (this will be reflected in higher client equity growth in the next 2.5 years).
The $800 million gross 2020 profit dollars in a hypothetical 0% interest rate environment has almost had zero benefits from this boost. We can safely boost the $800 million figure by 25% to get to a ‘normalized’ NIM gross profit of $1,000 million. For my annual profit estimates, I will spread the 25% boost over 2021 and 2022 allocating two-third and one-third, respectively:
Figure 12: Author’s model. NIM Gross Profit est. in zero rates environment (with one-off COVID-19 boost of 25% extra accounts smoothed over 2021 and 2022 growth numbers).
Commissions got a boost from market turmoil but are in secular decline as the core frequent trader base of “sophisticated individuals” (it is only logical that the most frequent traders join the broker with best all-in execution first) is diluted and IBKR client base grows.
Figure 13: Author’s graph based on IBKR monthly filings. DARTs per account
Q2 2020 saw the full benefit of this, with an additional higher gross profitability thanks to smaller order sizes, in general (typical in times of uncertainty).
For estimating underlying commissions, I simply use FY2019 commissions gross profit (which I define as commissions revenue minus “execution, clearing and distribution fees”) and grow them by predicted accounts growth of 15% and the one-off boost (as discussed previously) and correct for a secular decline (-3% CAGR over the whole period and slowing). This would suggest 12% sustainable growth, I model 10% instead, and the one-off boost from accounts growth.
Note, my underlying commissions estimate completely ignores the current elevated trading intensity (50% higher on a per account basis).
Most of IBKR opex is personnel (G&A and programmers) and 10% is occupancy, communications, and D&A costs (LSD growth). IBKR grew its personnel 7% CAGR in the last 5 years, and this was somewhat tempered by the sale of the market making business. I assume personnel will grow 10% p.a. for the brokerage business.
I will conservatively assume an overall long-term growth rate of 14% opex growth p.a.
My “underlying” numbers are not meant as reported earnings estimates. Why? I have not modeled:
- lumpy/cyclical elements (MtM P&L on IBKR forex, TIGR equity stake)
- increase in trading intensity in 2020 (temporary but will buoy profits a lot)
- one-off loss on oil contracts
It turns out reconciling my “underlying” numbers to consensus numbers is easy. Both higher interest rates in January & February ’20 (I modeled full-year zero rates impact) and higher trading intensity explain versus “normalized” explain the bulk of the difference.
Figure 14: Author’s earnings model.
The fall in interest rates has impacted IBKR’s net interest spread gross profit by 40%.
As NIM has client equity as the main long-term driver, IBKR’s fastest growing KPI (historically 29% CAGR), lower rates have pushed back IBKR’s inflection towards a higher profit growth rate by a few years. This one-time setback is magnified by IBKR’s continuous growth in opex (continuous long-term investment in more automation, products, exchanges, countries etc.) at LDD rates.
As shown in my underlying estimates, the high 71% of total gross profit in the mix from NIM will only be reached by 2023, resuming the sustainable >20% annual profit growth.
The beneficial effects from very high account growth between March ’20 and ~March ’21 (25% above the long-term trend) will only be felt in NIM gross profit dollars in ’21-’22 and completely offset the zero-interest rate environment impact.
I have used conservative estimates (key variables: supernormal accounts growth quickly reverting to historical norm by March ’21, 14% LT opex growth, 20% LT client equity growth, 10% LT CAGR in trading gross profit).
I argue the estimates show us that IBKR’s earnings growth formula is not impaired in a permanent zero rates environment: by 2025, IBKR’s profit will be double that of 2019, despite interest rates remaining at zero.
Rates at 1% in 2025 scenario
If the US rates turn out to be 1% instead of 0% in 2025, I estimate incremental gross profit to be $1.2 billion. This estimate is based on ~33% of rate hikes historically flowing through to NIM spread and also taking into account the heightened rates sensitivity between zero and 0.5% Fed Funds rate. For the latter, I rely on IBKR’s quarterly commentary on rate sensitivity.
This gross profit flows through 100% pre-tax and net earnings power would be $2.9 billion instead of $2.0 billion in a zero-rate environment. If ’25 rates were instead like ’19, net earnings would be $3.5 billion.
If rates would stay zero between 2020 and 2024, I estimate the total undiscounted net earnings impact from zero rates in this period would amount to $4.0 billion. If COVID-19 would not have happened; however, there would not have been a total boost of 25% new client accounts growth, which would make the ’25 estimate smaller by $0.83 billion. Capitalizing this extra earnings power in ’25 at a fwd. P/E of only 12X, the beneficial accounts effect from COVID-19 is large: $10 billion.
In other words, in this case, the pandemic (lower rates and more accounts) should be a net positive for IBKR shareholders as it pulls forward a lot of account growth, more than offsetting the one-off loss in earnings power from rates.
I have discussed a zero rates, a 1% and a 2019 scenario for 2025. Even before factoring in IBKR’s excess cash that has been piling up and is now 30% of the market cap, the P/E of IBKR in 2025 would be respectively 10X, 7X, or 5.8X. I leave the reader to discount excess cash which could be fully subtracted if IBKR were to be bought out by a larger bank or partially subtracted if IBKR pays another special dividend as it has done before in anticipation of higher tax rates. With any reasonable weight given to excess cash, the 2025 forward multiples are well into single-digit territory: much too low for a high quality business with double digit EPS growth.
Disclosure: I am/we are long IBKR. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.