|Shares Out. (in M):
|Market Cap (in $M):
|Net Debt (in $M):
|TEV (in $M):
Sign up for free guest access to view investment idea with a 45 days delay.
Fwd Numbers Above are Consensus Estimates
Long STT, one of the oldest financial institutions in the country and one of a handful of Tier 1 custody banks, because we believe it is too cheap at ten to eleven times (growing) EPS. That discount is derived from undue pessimism around pricing and a desire to group them in with a traditional large bank. In fact, STT is not a traditional lending-based bank but instead is a nicely defensive provider of crucial financial infrastructure and has the ability to grow earnings both with global asset growth as well as through improved operational efficiency.
2019 was a trough year for STT as it was dealing with temporarily increased fee pressure, management change, the loss of $1tn in Blackrock AUC as well as the challenges of integrating their Charles River acquisition all while interest rates continued to go lower. It was not a fun time for the company, as suggested by the almost 50% peak-to-trough decline in the stock. As a result, earnings dipped to $6.16/sh in 2019 from $7.29/sh in 2018.
Starting in 2H19, [the new CEO] Ronald O’Hanley’s initiatives resulted in fundamental improvement in the business. These initiatives were primarily around pricing discipline and cost controls. For many years, custody banking was an oligopoly that earned comfortably double-digit ROE’s due to rising global asset levels and meaningful frictional switching costs for customers. As a result, costs in the organization bloated with very little in the way of innovation or improvements to the business. Custody banking services represent a trivial part of overall costs for most asset managers, often measured in the single or even fractional BPs, and per-unit costs went down but that was materially outweighed by the sheer scope of the growth in custodied assets. However, with increased consolidation and fee compression in the asset management industry, the large asset managers (eg Blackrock, Vanguard, Fidelty, T Rowe) started to look closely at all costs - even the ones that were historically 2-3 bps. That led to a round of harder-than-normal price negotiations for custody banking services for equities in 2017 and 2018.
Servicing fees are a function of Assets under Custody/Admiration and have historically grown at a blended average of equity & fixed income assets. That was good for a ~5% 20-year CAGR. Due to the commodity nature of the business, STT, BK, and JPM roughly maintained the same market share as large asset managers doled out business roughly equally (legacy business almost never switches - so new funds/products are usually where they can create some negotiating leverage). Historically and currently there is very little observable direct price competition between the various custody banks.
Growth of Global Custody Assets, source: Globalcustody.net, AUC CAGR of 5%
Source: McKinsey, Revenue CAGR of 3%
Since it costs very little to service an extra unit of AUC/A, asset managers have always demanded fee breaks as assets grew. Historically this pricing pressure was around 2%/yr per unit of AUC.
Starting in 2017-2018, this fee pressure expanded to 4% with obviously high decrementals. Due to the nature of these contracts, there are 2-4 quarters of lag before one would start seeing the reduced revenues in the financials. Legacy sales incentives called for growing AUC (as opposed to business profitability) at all costs so bloated sales team at STT in high cost locations were happily complicit in this more aggressive set of pricing practices. Starting in 2019, STT escalated all pricing discussions to the management level which has allowed for pricing discipline and less middle-management sales personnel. The past two quarters we have seen fee pricing pressure stabilize (esp net of mix shift) and we would expect to see this continue to flow through over the next year-plus.
For us, that is one of the two crucial insights: we are not just predicting moderation on pricing, we are observing it. The other crucial insight is that the company is making good on its cost-cutting (more on that later).
Yet the stock still trades at 11x fwd Consensus PE despite guiding to 10% EPS growth in 2020. Several factors are contributing to this cheapness. In addition to the market inappropriately grouping this with traditional financials, the main recent cause appears to be the sentiment around historic comp BNY Mellon (BK) which is facing a whole slew of problems that STT has already encountered and addressed the past two years. This is partially because Fixed Income (where BK is stronger vs STT) has not yet gone through its round of hard pricing negotiation.
BK has re-rated materially lower and that has dragged down the sector. We would observe the following causes and facts
1. Charlie Schraf's departure to Wells
2. BK is early into the fee compression cycle – as their AUC mix is less equity. Fee pressure coming from other asset classes
3. Their expenses are up while STT’s are down. STT FY expenses were down 1% vs +1% BK. STT has more room to cut and we saw that in 4Q19 when expenses that quarter were down 9% YoY. STT expects expenses to be down 1% next year, BK expects them to be up 2% as the recent guidance.
4. Led by expense control STT’s pretax margin was 29.1% - up 100bps from 2018, whereas BK’s was 30.5% vs 32% a year ago
5. Increased fee pressure at BK which saw servicing fee revenue down .5% QoQ despite +3.5% AUC/A growth in 4Q19. Whereas at STT, AUC was up 4% QoQ and servicing fee was up 2% QoQ suggesting rapidly moderating pricing pressure
A look the company’s own guidance shows that a 10-15% EPS growth is easily achievable if equity markets hold up - and we think this is likely sandbagged, if anything, given new management’s incentive to break the cycle of disappointment with the Street.
On the revenue side, the strong start to the year, as well as an extra $1.1T of assets should make 3% servicing fee revenue growth easy to accomplish. NII is anyone's guess but - given the historically low rates globally - it’s hard to envision a scenario where NII goes down materially from 2019’s depressed levels.
This is the other leg of the thesis and it is simple. State Street was historically a very cozy place to work - the incentive has been to overstaff because clients appreciate ‘knowing whom to choke’ when they have questions about back office activity. Revenue per employee was always lower than peers as a result - Revenue/Employee was 9% higher at BK than at STT, for example - and a lot of the business was focused in high-cost areas like New York when that’s just not necessary in a digital world aside from certain things like sales functions. This continued until recent days: headcount between 2013-2019 grew at a ~+5% CAGR which management has highlighted was too high.
Management teams have promised to get religion for a long time, but O’Hanley appears to finally be following through. According to the data and color we are gathering, STT is reducing headcount at high cost locations by 2-3% a quarter. As a result, STT’s expense control is ahead of schedule: it appears they took out $400M of costs this year, ahead of the previous guidance of $350M. So far, the company has publicly cut 3400 jobs, mostly in high-cost locations to boot. Whether the company can continue to realize efficiency gains remains to be seen but years of high IT spend could make 2020 the year they finally bend the cost curve and realize greater operating leverage.
STT has been an aggressive buyer of its own shares over time and O’Hanley appears to be no exception, doing an additional $500mn worth of buybacks in 4Q19 alone (they also increased the dividend by another 11%).
The prior management team indicated that ~100% of net income would come back to shareholders in the form of the annual dividend and buybacks. At these prices, we believe that buybacks are more accretive but that’s not something we would rely on management getting significantly more aggressive.
On another note, it’s worth noting that STT’s CET 1 Capital stands at 11.9% and the overall Tier 1 Capital Ratio Stands at 14.7% which is probably non-trivially high and the prior mgmt team flagged it as another potential lever for return. O’Hanley may feel differently but they clearly have the scope to increase the buyback for a time if they are so inclined.
Abbreviated Business Background
“Move fast and break things..” – said no one ever at a custody bank.
STT revenue’s come from three sources:
1. Servicing assets under custody – This is usually a fee charged on the assets held at STT but is also based on the mix of assets, geographies, and transactions. This fee has been in the magnitude of 1-2bps of AUC.
2. NII – this is the interest income STT earns on assets held on their balance sheet. Key insight here is that NII is not purely a function of interest rates as STT’s customers are sophisticated financial institutions that won’t just let STT capture a NIM on their float without either deploying assets or demanding a higher share of the NII.
3. Asset Management (SSGA) – like a Blackrock or Vanguard. Most of the ~$2T is in passive ETF’s, which earn low management fees.
In addition to these core revenue streams STT also earns revenues from trading (mostly FX) & securities lending. The drivers for these revenue streams is active managers trading and volatility (hard to forecast, but cycle this will probably be stable)
So what exactly does a custodian do?
Quick Answer: A lot of unsexy but critical middle and back office tasks that are an integral part of the financial system to ensure that transactions are settled, accounted for, and continued to be accounted for. These tasks range from regulatory tasks, market micro-structure related tasks, fiduciary tasks, reporting/accounting, etc all the way to settlement. Their customers are mostly investment funds (eg mutual funds, pensions, ETFs and many other structures) where the commercial logic (or even regulatory permission) to bring many of these functions in-house is limited given the huge scale of custody banks.
From our research, none of these tasks seem difficult or complicated in isolation (in fact some of them are downright commoditized) but doing them accurately and quickly at massive scale is extremely difficult. Further, almost all of these tasks are essential/critical for the day-to-day operations of asset managers. The cost of ripping and replacing custodians and fund administrators is a huge headache. Instead, customers achieve diversification by directing new business such as new funds on the basis of which custodian they want to reward.
Because custody banks charge fees for these middle/back office tasks, they have a very different earnings profile than a traditionally understood ‘bank’ despite the name.