Description

One of the things religions got right is that important things must be preached regularly. I consider most things I wrote here to be just that: unoriginal but important. For the atheists in the church of IBKR, I marked the bits of self-proclaimed originality in yellow. Put your faith in stats: the originality and return of my write-ups correlate negatively (spurious stat as the profitable data points suffer unfortunately from small sample bias).

I look forward to any feedback from the board.

Sermon

IB is ~5x cheaper versus competition in terms of all-in cost of trading, deposits and margin loans for individuals, and 5-20x cheaper for institutionals. Despite this, it has a leading operating margin of >60% versus industry of ~30-50%. Why? Automation. >80% of 1200 employees are – and have always been - highly paid programmers. IB grew their robust IT system steadily and 99% organically.

IB has a unique automated infrastructure to execute and clear trading in 26 countries, 120 markets and 22 currencies. Account opening, transaction lifecycle, corporate actions are fully automated. Recent evidence shows a growing wedge in cost competitiveness in international markets, primarily driven by increasing regulation and scale.

The long-term strategy (flywheel of automating, scaling, passing on benefits to clients in terms of pricing, free new features, internal trade execution) that has always been executed on is now firing on all cylinders as IB listened to some of its five client segments that rose in scale recently and then making those features available to the benefit of all clients. For example, checking account capabilities for financial advisors, improved capabilities in performance reporting and FDIC sweep for institutions, continuously improving and best-in-class API capabilities for prop shops and trade execution algorithms for institutions (the latter are old examples, but being improved upon).

IB currently trades at 22X P/E (excl. excess capital 16.4X). While its overall profit growth is not stellar, I’d bet on the longevity of growth. The human tendency to over-extrapolate longevity of growth that proves to be elusive is incidentally the main reason for the value factor outperformance. His argument better be good, I hear you thinking. My variant view:

IB’s growth rate will always be tempered by switching costs that are high in this industry. I believe the positive of this “negative” that is much mentioned by IBKR atheists is longevity of above average growth: IB’s value prop delta with competition is high for many but might not exceed the cost to switch. The result is the moderate growth rates we see. What we don’t see is the positive side: because new clients are born (or start investing) every day because of slow but certain demographics and worldwide middle-class (in IB’s case especially Chinese) getting richer, these do not bear switching costs when choosing their first broker. Demographics + absence of switching costs make me comfortable in calling for a long growth period beyond the realm of “consensus numbers”. In a nutshell, IB’s huge value prop is masked by brokerage switching costs for growth in existing brokerage clients but not in first-time clients. As a last thought, I will leave you to think about the lifetime value per reported IBKR client equity dollar from existing clients (older, richer) versus LTV/equity of first-time clients (younger, asset-poorer, income-rich). Why should I use LTV you ask? 1) Switching costs 2) IBs very low churn (Peterffy hinted to about ~1% per annum to competitors).

Why does the opportunity exist / Why now?

An inflection within an inflection, two levels?

The terminal decline of IBKR’s market maker business and the rare IPO-C corporate structure mask the spectacular historical growth of the brokerage unit in quantitative screens.

The following two conventional wisdoms about IBKR are rapidly becoming obsolete because of rapid equity growth, rising fed rates and stellar introducing broker “i-broker” equity growth:

• IBKR growth visibility is bad because of commissions: being the cheapest, IBKR first attracted its legacy most active traders. This base is continually being diluted by less frequent clients (types). Still, even on an average basis today, IBKR client base is widely known to be the most active by far. Two scares surround commissions:

  • trades per client will keep on falling (true)

  • the race to zero in commissions (overdone in IB’s case because of its exposure to institutional and non-US business)

  • Today, the mental shortcut that IBKR is most vulnerable to commissions concerns because it has the most active traders has become largely irrelevant. IBKR will earn almost 70% of its 2018 operating income from net interest income (this was 50% as recently as 2015). While client equity (a long-term proxy for both client cash and loans) has CAGR’d at 30%, trades have CAGR’d at only 10%. Rising US rates have drastically accelerated this inflection.

• IBKR does not benefit from higher rates: IBKR pays its clients interest for deposits at the benchmark rate minus a spread of 1.5% and smaller. IBKR only pays interest on clients cash on the portion above 10 kUSD equivalents (if any; per currency separately) only if net client equity is above 100kUSD. Interestingly, IBKR widely touts that i-brokers get discounts based on aggregating all underlying clients (i.e. looking at an i-brokers accounts as one big client). What is rarely mentioned however is that this rule does not apply for deposit and margin rates. While the i-broker can mark down the deposit rates to zero and mark-up the loan rates, the i-broker will itself still pay IB’s rates on an individual accounts basis. Crucially, the bulk of these underlying i-broker accounts are small accounts, allowing IB to earn the full fed rates on these clients. I-broker equity growth in 2016 and 2017 was resp. 53% and 94%. As a result, today I estimate that 33% of USD client deposits earns 0% (= the fed funds beta for IB). As i-brokers outpace overall growth, the interest spread and deposit beta for IB will continue to rise.

An inflection within an inflection within an inflection, three levels?

To summarize, as brokerage income emerged from the decline of the market maker, future brokerage income growth is now inflecting towards faster growth from net interest income (which itself is increasingly oriented towards higher margin i-broker free deposits).

Some other why-now’s:

Benefits from YE17 sale of market maker

• Closing some important gaps with Schwab/Fidelity: Checking account capabilities (credit card, bill pay, direct deposit). I speculate that Peterffy will soon announce differentiating cash backs on the credit card payments, already featuring the cheapest borrowing rates in the world (and EU/Canada added in ‘18)

• Vastly better UI on web and mobile. TWS on desktop to follow?

• Barron’s took notice: 2018 top broker

• Increased management focus on broker. Evidence:

• Wrong perception of trading against own brokerage clients buried once and for all

• 1.5BUSD of capital freed up

Overrated bearish perception of impact of “race to zero” in US commissions. IB has almost nothing to lose:

• of those 25% of US individual clients, many either are

• educated and understand importance of execution quality

• at IB for the unrivalled differentiating availability and cheapness of international investing

• only ~25% of clients exposed: only half of clients are US based, while only ~half of those are individuals that tend to care more about headline commissions (vs institutionals caring more about execution quality, deposit & margin loan costs, short borrow)

Growing evidence IB has a unique asset in its fully-automated international execution and clearing capabilities, confirmed by

• nearest competitor TradeStation (from its point of view) becoming an IB i-broker in 2018 for its international brokerage business

• surging growth in introducing broker segment (+33% account growth YoY ’16 and +61% YoY ’17) primarily driven by Asia (3 years ago, Tiger Brokers was founded. Today I estimate it represents 2% of IB’s trading volume) confirming its “platform” appearance

• Scottrade in 2015 (though AMTD pulled that plug after acquiring it)

• E-Trade left the international brokerage business altogether

Value drivers: inflection from commissions growth to equity growth

The usual numbers.

While brokerage growth at the group level has been masked by the epic decline of the market maker (80% of profit in ’08, 0% now) – not to mention negative distortion of top-line per share metrics and fake dilution because of the up-C corporate structure - brokerage revenue CAGR itself has been masked by slower DART CAGR (10%) because, historically, commissions were more dominant source of revenue (in part because of low rates). Due to a confluence of much higher growth in equity and higher rates, 70% of profit in 2018 will be from net interest margin “NIM” that is primarily driven by client equity growth. Have a look at the above table to contrast the two relevant KPI’s for the past value driver and the future one.

How is NIM earned? In one formula = Client equity x Sumproduct (interest spreads; client allocations % of equity) sum over deposits and loans. Let’s back up the ingredients we’ll have to tackle:

• Client equity (LT driver)

• Interest spreads (cyclical, some LT mix effects)

• Cyclical because of spreads are impacted by general rate levels

• Mix effects because of higher spreads on small vs large clients (i-broker growth is expanding IB’s spread)

• Client % equity allocations to interest generating activities deposits, loans (cyclical)

  • Make margin loans (50% gross interest revenue)

  • Invest in short-term paper (30% gross interest revenue)

• IB pays clients for their gross cash positions to (-20% gross interest revenue)

• IB is intermediary between clients/institutions and other clients/institutions borrowing/lending securities for short-selling (20% gross interest revenue)

Knowing we need to keep an eye on NIM, let’s have a look at the client segments and their TAMs.

Figure 2 Peterffy believes penetration in Individuals should "eventually" go to 100%, while prop shops growth is largely behind us.

I marked the segments with high market share in red and orange. A quick look confirms that the greenfield segments dominate most in terms of ... the most important value driver. For all other information on client segments, I really recommend the last IBKR post.

Various sources of anxiety-inducing short-term noise

Part of the short-term value drivers for IB are cyclical behaviour of clients in terms of

• DARTs (cyclicals ignored because rather at cyclical low & secular decline is modelled)

• Client allocations as a % of their equity to margin loans, cash

• I model a very fast reversion to “normal” levels by January 2019 causing a drag on growth in ‘19. For transparency’s sake, my assumptions are visualized below.

With client % allocation of equity out of the way, we are left to interest rate spreads.

Forecasting those spreads calls for forecasting two things:

• avg margin loan spreads IB charges above benchmark rates, impacted by client mix (on the deposit side, everyone eligible gets the same benchmark - 0.5%, but on the margin loan side, the large borrowers pay +0.3% while small borrowers pay 1.5%)
• benchmark rates themselves, as IB does not pay interest on ineligible deposit balances (<10kUSD equivalent of currency OR in accounts smaller than <100kUSD) making the spread somewhat sensitive to the benchmark themselves. I estimate about ~1/3rd of cash deposits to be ineligible balances. For the great bulk of eligible balances, the spread is only constant unless rates dip below 0.5%

Forecasting average margin loan spreads: mixed

IB recently saw a surge in margin loans especially from hedge funds (small and large funds alike. Large funds tend to have an n-th account with IB to cherry-pick certain capabilities). In fact, the smallest spread charged used to be +0.25% but Mr. Peterffy did not want to source margin loans from outside (typically exclusively sourced from internal customer funds) and put the brakes on by raising to +0.3%.

The only client type that is growing much faster than others is introducing brokers. IB advertises to treat all intro-broker underlying accounts as “one account” in terms of pricing. I was surprised to learn that for margin loans and deposits, however, IB treats each underlying intro broker account individually.

Because intro broker accounts tend to be much smaller in account size (the avg account size of IB is at 250 KUSD ~Schwab & Fidelity, while most online discount brokers with less sophisticated customers have average account size of ~20-60k. What this effectively means is that IB is growing very fast in ineligible deposits (paying 0%) and the relatively higher spread margin loans.

Last year, client equity growth at introducing brokers was +94%, up from +53% in ’16. Account growth was +66% and +33% respectively. Because new individual accounts typically double in size after 18 months of signing up, a part of the +66% account momentum has yet to trickle through to equity growth. The funding cost for IB is improving as the % of ineligible

For simplicity I model no mix changes in spreads.

Benchmark rates

• What about rates rising? Isn’t this the #1 theme for financials, so it’s happening right? Other brokers are more interesting in that case!

• What about the market crashing and rates crashing with them? It’s overdue!

I encourage everyone to open an IB account to express their strong views in one of the world’s most liquid markets. It is true that very close to 50% of these trades make money at IB after fees! Ahh, the temptation to forecast Howard Marks’ “important and unknowable”. I will skip the preaching for the church of the efficient market on the topic of interest rates, I am a sinner too.

I model no further rate benefit after Q2 2018.

Model

To summarize, the model uses:

• Q2 2018 Fed rates going forward (not baked in in ’17 financials, hence source of ’18 growth)
• no change in NIM spreads despite favourable dynamics at intro brokers accounts
• a quick reversion of client % of equity allocation in margin loans (and to a lesser extent cash) to more normalized historical levels by January ’19 which impacts the model mostly in ’19 resp. negatively and positively
• most importantly, 19% client equity growth from ’20 onwards and 7% commissions growth. This drives an increasing overall growth rate by 2022, i.e. ~14.5% revenue and 18% profit growth. See below for evidence that equity growth has rarely been below 20% p.a. for IB

Interesting to note is that equity growth delta with S&P returns was particularly high in the wake of the financial crisis. Mr. Peterffy noted at the time that clients increasingly think about costs and safety in these times (IBKR ran a WSJ ad featuring the word “conservative” in a dictionary template).

After the recent sell-off, IBKR is selling at 22x P/E.

My numbers are roughly ~ consensus until ‘20. Where’s the edge?

Variant view

Longevity of growth beyond: IB’s growth rate will always be tempered by switching costs that are high in this industry. I believe the positive of this “negative” that is much mentioned by IBKR atheists is:

• Longevity of above avg growth: IB’s value prop delta with competition is high for many but might not exceed the cost to switch. The result is the growth rates we see. What we don’t see is the positive side: because new clients are born every day (slow but certain demographics & Asian macro), these do not bear switching costs when choosing a new broker. Demographics + absence of switching costs make me comfortable in calling for a long growth period beyond the realm of “consensus numbers”

• Barrier to entry: who is going to try beat IB at being even slightly lower cost if the lead time to profitability is so long?

Conviction in those consensus numbers

• In most industries, growth revenue typically comes with more risks than existing revenues. IB is highly scalable

• I believe underlying assumptions of consensus are quite conservative when comparing to IB’s history (30% equity growth last 10 year) and recent evidence from product improvements

More thoughts on longevity of growth

• 70% of 2018 income will be net interest income “NIM”. >55% of current client equity – the most important KPI driving NIM – is from greenfield client segments (<0.5% of TAM penetrated in fields of global retail clients through intro brokers, financial advisors and hedge funds)
• Equity growth was recently converging to account growth (see figure)

  • while this is because of mix shift (higher account growth in intro broker underlying accounts that tend to be smaller retail investors), mix shift effects fade away eventually

  • positive corollary to newer smaller accounts: the brokerage industry has high switching costs for clients. The most active individuals and prop. shops already joined (i.e. switched) more than a decade ago because the delta in value prop was so large versus the switching cost and Peterffy estimates to have penetrated 1/6 of sophisticated individuals. Today, growth is increasingly from new/affluent/young investors. From a customer lifetime value perspective (IB’s churn is very roughly ~6% and only ~1/6 of that is switchers), would you rather have new account dollars being reported from 25-year-olds, or 55-year-olds? Indeed, these young customers have a higher income to asset ratio, hence potentially drive much higher equity-per-account growth that should lag the high current intro broker account growth

“Most people in Europe already have an account and most people in the United States already have a broker. In Asia, it is the newly rich, the young people who open a brokerage account for the first time in their lives. So they compare the brokers. As anyone who compares brokers can blindly see that we are by far the lowest cost. So, it's not that the message is more, it's stickier, it's more like the audience is different. And that is the large picture.” – Peterffy Q1 2016