We are struggling with the question of how do we compete, and one answer is to look at efficiently deploying capital. So we have invested a lot of effort into building both quantitative and qualitative models across the banking sector, in order to evaluate optimal capital.
Mike Ensweiler:Welcome everyone to the sixth episode of In Your Best Interest: An ALM First podcast. A show that will explore common depository challenges, give you an insider's view of the latest market trends, and share stories and insights from industry leaders. I'm your host, Mike Ensweiler. Economic uncertainty, cybersecurity threats, non traditional non-depository lenders, excess liquidity, compress margins, threats of defaults. What does this mean for our balance sheet? Do I have enough capital to weather the storms? How do I stress test these scenarios? This week's episode will focus on capital planning and stress testing in this current environment. Today, we are joined by Macie Fults and Kevin Kirksey. Kevin is a principal at ALM First joining the firm in 2011. Kevin oversees the Merger Advisory team, our Strategic Solutions Group, which conducts merger valuations, capital planning, and stress testing ALM and CPST model validations as well as CECL analyses. And Kevin also runs our operation and technology groups. Macie Fults joined ALM First in 2018 as an associate for the Strategic Solutions Group, Macie performs capital stress testing, ALM model validations and CECL valuations. Thank you so much Macie and Kevin for joining us today.
Kevin Kirksey:Thank you, Mike. It's my genuine pleasure to be here with Macie and I'm grateful to all of the listeners engaging as we discuss the new normal and capital planning and stress testing.
Macie Fults:Thank you so much for having us on to talk.
Mike Ensweiler:Well, you know, before we get to the theme of the day, which you mentioned, Kevin is his capital planning kind of in this new norm, you know, I'd love to engage the audience and have them learn a little bit about the people on the other end of the line here. And so maybe just tell us a little bit about yourself outside of what I read in the bio.
Kevin Kirksey:Yeah, absolutely. Um, I am very involved with a lot of nonprofits in the city and, uh, sit on several boards. And the one that I'm most proud of is a rescue group, Dallas Pets Alive, um, where we do not deny any animals based on behavioral or medical conditions.
Mike Ensweiler:That's cool. Well, that's gotta be a pretty fulfilling. And how about you Macie?
Macie Fults:Yeah. Um, kind of what I do for volunteering typically towards this kind of part of the year is I am on the Rice alumni volunteers for admissions. So, um, I spend the last kind of like six months of the year interviewing for, um, undergraduate programs at, uh, Rice.
Mike Ensweiler:Awesome. Well, that's pretty neat. So Kevin and Macie again, you know, just maybe tell us about the path that led you to ALM First. Well, we'll start with you, Kevin.
Kevin Kirksey:I was always interested in learning how companies operated and I had a unique skill set in terms of applied mathematics that led me down a path of becoming very interested in financial institutions. And I was really excited for the opportunity to work at a company that had a niche presence, um, that focused on a lot of institutions that really did have a mission orientation and, um, serve the underbanked.
Mike Ensweiler:How about you Macie?
Macie Fults:Yeah, so, uh, I graduated from undergrad in 2017 and my first job out of college was, uh, working in program finance for a defense contractor and that was fun work. I learned a whole lot of things. I got really sharpened by soft skills, but it wasn't necessarily as quantitative as what I was looking for. Um, my undergrad was in mathematical economic analysis and I really wanted to be able to kind of combine the soft skill set and some of those more firms skills I had on the analytic side. So I reached out to a recruiter and I started looking for kind of more quantitative, potentially consulting type roles to marry those two worlds. And I laid it all in first. It's been really great. It's exactly what I was looking for.
Mike Ensweiler:Well, and the timing probably couldn't have been better or you had an opportunity to sharpen your skills before we got into a pandemic and talks of modeling negative interest rates and all kinds of things that I think will keep your, uh, your skills sharp for years to come.
Macie Fults:Yeah, definitely. It was, it was kismet.
Mike Ensweiler:You know, there's been a lot of discussion on the appropriate amounts of capital. We've heard from some of our other guests on podcasts state that the industry was overcapitalized coming into the current crisis, this pandemic, the crazy 2020 that we're living in now. So how do I, as a balance sheet manager, know how much capital is enough capital. And I'll, I'll throw that out to both of you.
Kevin Kirksey:Mike, thanks for the question because we're all rapidly wrestling with this pandemics global scope and in an enormously volatile election year with ultra low rates and that's why you mentioned earlier, some people are even modeling negative rates. We've seen a lot of organizations re-baselining their projections and using a lot of the Feds, uh, June, 2020 alternative scenarios to do so because they are asking the question of, are we over capitalized? And historically we had that knee jerk reaction after the great recession of 2008-2009, where we saw capital requirements for financial institutions increase globally. And we became at a point where we weren't as efficiently deploying capital, um, at a systemic level. And when we look at competition today, whether that's We-Chats deposit capabilities, shadow banks, neobanks like Chime and Viro, we are struggling with the question of how do we compete. And one answer is to look at efficiently deploying capital. So we have invested a lot of effort into building both quantitative and qualitative models across the banking sector, uh, in order to evaluate optimal capital levels. And that is something that has to be married with your board's governance and risk tolerance. And that appetite is really what's going to drive the ultimate number. So it's certainly not a one size fits all answer. So I've discussed the motivation for answering the optimal number of capital and some of the key considerations we'll be discussing on this podcast. I'm sure some more tactical measures that can be taken to arrive at that answer.
Macie Fults:I would like to add to that is whenever we talk about, um, strategically aligned appropriate amount of capital, and we had a webinar on this recently, um, like Kevin said, there's a whole lot of little micro, uh, decisions and components that go into that. But I like to think of it as three main parts. And that is you have your regulatory and board required minimums that you must maintain. Uh, you need a good buffer to be able to absorb potential losses, um, and take into account your exposure, should they occur. And then also you want to have enough leftover to fund your strategic initiatives. So making sure that you're covering both your need to do's and then your want to do use. Um, but as Kevin said that number's going to be different for everyone based on what their initiatives are.
Mike Ensweiler:Yeah. That makes perfect sense to me. There's something Kevin, that you said that I'd like to just maybe touch on because it was really interesting and I'd like to just go a little deeper on it. You said kind of pre pandemic that the industry as a whole wasn't efficiently deploying capital on a systemic level and that to compete going forward, you're going to really need to evaluate what those optimal capital levels are. Can you just kinda explain what efficiently deploying at a systemic level? What did you mean when you said that?
Kevin Kirksey:Yeah. So if we look at large, large figures across the industry, as a whole, you can observe clear trends that there are high amounts of liquidity being held right now. And that exemplifies that when we had payroll protection program come in, when we had this flight to quality that we've seen across the entire ecosystem while sure nobody was able to predict the pandemic. And if you did give me a call, I'll make it crystal ball of yours. But it's very evident that everyone became in triage mode deciding what do they do with this excess liquidity? And if we were in a fast acting efficient deployment of capital system, we would not have seen these large levels of liquidity being held for such long periods since the pandemic began earlier in March, that really affected a lot of banking institutions. And so if we think about how those competitors that I mentioned earlier, these technology focused firms that don't have bank capital requirements, they're already doing things to harness natural language processing, machine learning, and they are the epitome of agility and automation. And if we had our capital planning as automated as those firms structures are, then we would be regularly turning out a very efficient and automated capital flow.
Mike Ensweiler:And that sounds like it would be a pretty heavy lift across the industry.
Kevin Kirksey:Extremely. And that's why right now, um, collaboration is more critical than ever. Um, and we're hoping that today's talk will galvanize further peer dialogue, uh, so that everybody can optimize, uh, their capital positions and make that deployment process as efficiently as possible because across the entire ecosystem, it is going to be an extremely heavy lift to truly future- proof these capital positions.
Mike Ensweiler:Makes sense. Talk to us a little bit about the differences between capital planning, kind of, you know, kind of the day to day, year to year capital planning I should be doing. And the major tests that most of us have heard about namely DFAST and CCAR, what does capital planning mean or entail and how is it different than the federal reserve stress?
Macie Fults:So when we talk about a DFAST and CCAR, those are, um, regulatory requirements and, uh, you know, there's differences between the two, CCAR is more involved, but, uh, the main kind of difference is you only need to perform DFAST. Once you hit 10 billion in total assets and in CCAR is four 50 billion plus in assets. Um, those are regulatory driven exercises and they're very useful and they have practical implications. But at the end of the day, if you're trying to meet the exact rules of say DFAST exercise, it's because you have a regulatory requirement to do so. Capital planning from just a day to day internal operational standpoint is something that is not required, um, obviously by regulators, but you want to be doing if you're managing capital. So that's doing things like looking at your risk tolerances, uh, reaching out to different departments and saying, what are our material risks? And, you know, there's two parts to that. Uh, what's the probability or the likelihood that we would face, which risks and what are our actions going to be, um, on the other end of that, given our exposure at occurrence. Um, so it's important to have in your policies and make sure that when you're doing capital planning, you're aligning it with your risk appetite statement with your risk framework, so that you have a kind of cohesive, uh, narrative inside the institution so that if and when, uh, adverse things happen, or even if good things happen, you know, what you're going to do in a time of crisis or in a time of, you know, um, windfall.
Mike Ensweiler:Any additional thoughts on that, Kevin?
Kevin Kirksey:I think that was a good summary. I think that today, one of the big imperatives to focus on and differentiate between these various tests is looking at qualitative overlays. So no matter if you are unregulated from capital planning and stress testing, your DFAST or CCAR, you want to look very closely at that qualitative overlays because the landscape has changed irrevocably, um, specifically in terms of acceptable forecast accuracy. So while institutions are relaxing goodness of fit expectations, and they're more backtesting on this recent experience during the pandemic, um, that has forced the industry to look at things in their risk taxonomy in new ways. And that's really outside of the regulatory stress testing framework. So to give simple examples of that, I would look at new cyber threats posed by a majority of the workforce working from home, um, at, at your bank or your credit, looking at the difference in scenario design, from an epidemic perspective between Manhattan and South Dakota, looking at operational risk diagnostics for e-signature or digital lending applications, and what impact that has on your expected new volume growth and what that has on any operational risk, where you need to have additional capital buffers because of these new initiatives. So irrespective of the regulatory requirements, these are the types of qualitative questions that we're seeing being faced in the industry today.
Mike Ensweiler:And let's stick to that point. You know, there's been a lot of discussion around modeling negative interest rates that COVID-19 credit concerns, you know, namely increased defaults, as you mentioned, some kind of cybersecurity threat or cybersecurity failures, natural disasters, and we've got two hurricanes storming towards the Gulf coast right now. So how should I, as a balance sheet manager be thinking about these risks as it relates to my capital plan? Can you take a little deeper dive on that?
Kevin Kirksey:Yeah, absolutely. And what obviously complicates this, uh, for US institutions is, uh, our, our fiscal policy and central bank stimulus plans, uh, that we didn't have access to really during the great depression or weren't deployed. So at a micro level, we're going to be facing unique challenges as we try and predict how these, uh, various stimuli are going to impact, uh, micro risks. So we saw that recently with student loan borrowing programs, um, extending forbearance and, and these types of challenges really, um, hamper models. So that's why my previous point about really looking at these qualitative overlays, it's really shifted thinking about model risk and model dynamics. So to answer your question more specifically, what can be most advantageous right now is if you have those three lines of defense in house, perfectly lubricated you're, you're accessing your exposures on very granular levels. And that's critical because a lot of conversations, we're a part of we're hearing boards and staff say we have too many, uh, investments relative to assets. We can't loan them out. We have all of these deposits and we don't have analytics to determine how long they're going to stay here. And this is really where everyone starts rustling. And instead of looking at balance sheet mix as a whole or asset mix, start breaking down each individual portfolio into unique risk tranches, um, to determine if this is a business line that you qualitatively want to scale rapidly, and then look for effective challenge against that business decision. So it's kind of the reverse instead of building out a bottom up model, um, people are starting to focus more on those qualitative type decisions in the boardroom because no model was really able to predict with much accuracy, uh, the impact and the, and the length of this pandemic. Anything to add to that Macie.
Macie Fults:Yeah. Um, all I would add is when you are, as you mentioned, there's several different kind of idiosyncratic different risks that you can evaluate. I would say that definitely put the work up front and seeing which risks are actually material to your institution because, um, different geographies, different member bases, there's all kinds of things that can go into that. So I wouldn't, uh, it's more about your quality of your risk assessment than your quantity. So the, rather than trying to run, you know, all 10 that you can think of, uh, really talk to your department heads, say, what are our material risks and focus on making those risks scenarios, um, as meaningful and as helpful as they can be. Again, forecasts are almost never, right. But if you're focusing on the things that do pose a threat to you, and you're really putting the time and effort into adjusting your assumptions, um, you're going to be able to develop actual, useful strategies on the other end of it. And that should always be the end goal.
Speaker 3:Yeah, Macy, that's an excellent point because we've seen a lot of lenders that had exposures to hotel restaurant, office commercial real estate, and their variable selection process while respectful in terms of finding statistically significant outcomes, pre COVID-19 environment, um, the adjustments that they needed to make and read back to us, um, were, extreme. And so if we extrapolate that further and you talk about institutions trying to grow organically during this time. So instead of being defensive, if they are trying to migrate down the credit spectrum or increase non-QM position mortgage loans, or start looking at some more esoteric loan asset classes like equipment lease or vacation ownership interest. Their model risk team is going to be very, very busy for a long period of time trying to get under the hood and their arms around all of this. So I think that these, these types of vulnerabilities that are being exposed now of marrying qualitative overlay with the numerical analysis makes the board's job of looking at a model inventory with decision trees, uh, really come to light and they can say, is this a model where we're going to trust our clients to tell us what the risk exposure is, or is this something where we need to have the effective challenge as business strategists that are governing the organization? And I think we're leading more towards the ladder.
Mike Ensweiler:So let me ask you this, you know, you terms, model risk teams and those kinds of things. We all talk to depositories that are you know, hundreds of millions of dollars in assets are smaller. They don't have model risk teams. They may not even have a model in house. So, you know, kind of breaking it down for the vast majority of institutions out there, as I think about this, assuming I'm in that institution and I'm part of the ALCO or the treasury team, how should I given my size, um, and be thinking about stress testing, my balance sheet?
Kevin Kirksey:Yeah, I think the first thing you can look at are the publicly available results from the June 2020 alternative scenarios from the Fed. And you can take those high level portfolio segment loss rates and apply that to your portfolio and see what that does to your capital position. And that's just quick multiplication in Excel, uh, does not require any form of large scale model. And, uh, this is something that institutions of all sizes. When I talked about looking at unique vulnerabilities, you may not have a full fledged model risk team in house, but you can at least have one person spend a few hours going through your TDR portfolio and looking at those high risk segments and the smaller you are, the fewer loans it will likely be. And you may be able to do kind of an individual check. Um, some smaller institutions may know exactly who the borrower is, and they may able to determine how much risk exists from a credit perspective. Uh, other things that can be done are looking at the metrics that, uh, your boards evaluate and making sure that you're providing them, uh, helpful information and helpful data. So you may not be able to codify every decision that is made inside of your organization, um, when you have less than a hundred million in assets, but you can certainly be thoughtful of what metrics you're sharing. So an example would be if the board is concerned about the retail deposit portfolio and that excess liquidity show them what percent of those are from direct deposit relationships? What percent of those have an uninsured, um, dollar amount? What percent of those came in and transact primarily through digital channels? And then possibly looking at things like what is the general demographic profile in terms of employment base and age for that deposit portfolio. And those are things that can be presented even at a very high level for any size institution.
Mike Ensweiler:You both talked to a number of institutions, worked with a number of institutions in the area of capital planning, as well as the stress testing and the DFAST and so forth. What are some of the adverse or severely adverse stress scenarios you've been discussing with clients in the current environment?
Macie Fults:Yeah, of course. So, um, obviously you have your, uh, Fed/NCUA supervisory stress test scenario. So we always look at that severely adverse scenario. Um, obviously that's, as we said, forecasts are not, uh, accurate more often than not in those don't claim to be true economic forecasts, but we've seen the market do different things, the environment for the last eight months or so is not, uh, what you would have in a severely adverse scenario. So we have had a lot of conversations recently about COVID-19 scenarios. Now I say scenarios because we still don't know what's going to happen. Um, it's a new thing every other day and, you know, is it going to be a V-shaped recovery? Is it going to be U-shaped? Is it going to be a W-shaped? You know, are we going to go up and down for the foreseeable future? So that's the main one that we've kind of had interest in as of late. Um, and, you know, we like to approach that from a few different angles because we don't know what's going to happen. Um, and that goes into, you know, what our collateral prices going to do. What's the volatility going to do, um, unemployment. So particularly, do you have a lot of members or, uh, you know, your deposits and your lending tied to certain industries that are going to be more adversely affected than others. Um, and another kind of really important one that more and more clients are having conversations about is the cyber security scenario. So what would be our impact if there is a breach, uh, how would we address it? What would it cost us to address it? And it's important to also remember that when we talk about things like cyber security, aside from just your initial loss, um, and fraud, or, you know, heaven forbid embezzlement, anything like that, there's also reputational risk on the other side of that. Um, if something adverse happens like that, you have to worry, okay, we handle the initial fire, but then there's going to be lots of little fires because, uh, you know, when it gets out to our members or to other people in the public who may be, we're trying recruit as members, um, how's that going to tarnish our reputation and how do we recover? So those are kind of the, uh, the ones that stand out to me as the most relevant that I've been having conversations about.
Mike Ensweiler:That makes sense, how do I incorporate capital planning and stress testing into my strategic and operating plans?
Kevin Kirksey:I think the first step is ensuring that the capital planning process has effective challenge, not just across business lines, but upstream with the board of directors as well. So integrating capital planning and stress testing, um, becomes an enterprise wide effort. And in order to assimilate some of the best practices that we've seen in the industry in terms of using stress testing results, to inform the capital plan, what's different about that recommendation today than it was nine months ago, is that there's going to be a lot more sensitivity testing done on the actual stress test results because of the lack of predictive power we've seen with this pandemic. And then there's also going to be some sensitivity analysis done on the plan itself and determining if there is an alternative path to explore from business strategy. And the only way to achieve that level of integration into strategic planning is to involve all of these, these key stakeholders in the discussions and not have this led by just a handful of executives. There's gotta be buying in from numerous parties.
Mike Ensweiler:So that's a great point. Kevin, having kind of all the functional areas be part of this. So talent sounds to me like this, isn't something that you kind of knock out one week. This is, this is a process. So how long does it take to build an effective capital plan? What does a process look like?
Macie Fults:So, um, obviously this is a, uh, kind of complex engagement, a little nebulous. There's a lot of things that go into this. Another important part of capital planning is that it, it really is all hands on deck. So you want to make sure that you have ample time to meet and consult with all departments and, uh, make sure that everyone is kind of giving their input for that reason. It can, it can absolutely take, uh, years to get your capital plan exactly where you want it. If you're approaching the 10 billion mark and you, you're going to be beholden to those DFAST requirements, we recommend getting started on your capital plan and stress testing, kind of a run-throughs early, uh, before you hit, before you cross 10 billion, it's going to take one to two years to get a turnkey model and easily updatable capital plan, uh, functioning. So time is of the essence. You definitely don't want to wait until you cross a certain threshold and then be scrambling to put it together. With that being said, your year one capital plan is not going to look like your, your ten capital plan, but that doesn't mean that the first year doesn't hold value. So even if you don't have a full team that can dedicate 40 hours a week every week to this, um, it's definitely worth getting started where you are now and putting it together. And what I would say is, you know, six months into a capital plan, uh, that six month product is going to be better than your, I never even started a product. So rather than worrying about too much, uh, this is a huge undertaking. You know, it's going to take years. We need to wait until we have X amount of staff, staff for X amount of time, uh, just get started and, you know, take it as slowly as you need to, and you'll end up in a much better place than you would have otherwise.
Kevin Kirksey:Yeah Macie, the idea of just starting is the most important. And if you are an institution that has limited data today, then start harvesting more information from your core and from your sub-servicers start capturing more data at origination. Those are things that every institution can can implement. Then from there, the next step would be, um, benchmarking implied loss rates, uh, under the various scenarios that are published by the fed. So not just the alternative, uh, 20/20 scenarios, but just looking at the, the DFAST results that are published. And you can still ascertain some analytical insights from those against your own portfolio and with adequate effective challenge and proper documentation, you can be in a better place than you otherwise would. Um, with just some caveman level math, you can also start to be evaluating your model inventories in determining what are the most critical models that you want to develop. And not everything has to have a model on day one. Some of this is going to be done over time. So you've heard me talk a lot about qualitative overlays, and this is an effective strategy as long as there is sufficient documentation.
Mike Ensweiler:So it sounds like to get started, um, asset sizes, it nearly as important unless I'm approaching a$10 billion threshold, but this is something that should be part of my regular business practice. And I don't need high tech, you know, expensive models and teams to start this process. I think you gave us some really good ideas of a couple of things we can start thinking about and doing ourselves in house. And at what point does it make sense to bring in some additional expertise?
Kevin Kirksey:It comes down to intended use of capital planning and stress testing results. So for the smaller organizations, they may just need to evaluate at a high level and looking at their information in aggregate could be perfectly sufficient for their risk tolerance and their business strategy. It really comes down, it's not so much a question of size as it is a complexity and decision making framework. So if there is an organization that wants to enhance their risk based loan pricing, bringing in a third party that has access to a larger quantitative tools and more experienced teams can benefit from that arrangement as opposed to trying to build and acquire all of that talent and systems in house. But there does become a point on the flip side where due to speed of decision making, you may be willing to spend significantly more to have that talent and those systems in house so that you can rapidly control the process. And so that's kind of the life cycle or an evolution of CPST for organizations.
Mike Ensweiler:You know, we're just starting to scratch the surface and sadly we're budding up against the clock. So, um, maybe you can just leave us each of you and I'll start with, you Macie, just some, some closing thoughts. What are some things that, um, we should leave people with, um, that we didn't talk about to this point?
Macie Fults:Yeah, of course. So, uh, I think the overall kind of theme that I would like people to take away from this is that, uh, capital planning and to some degree stress testing in general, it's not only for, uh, larger institutions. Truly, uh it's for anyone who manages assets, uh, you know, credit unions in particular who have, um, these member mission statements where they're trying to, um, achieve these positive things in their community. It's worth taking the time to sit down and really start flushing out your capital plan and really evaluating your risks and just making the baby steps towards it, because it's ultimately leads to a much better product than in to the end. And you're also to get a lot more out of it. If you don't wait until someone tells you, you have to do it. So it's a positive thing. It's a good thing. Definitely something to start thinking about it any size.\.
Kevin Kirksey:I couldn't say that any better. What I would add is that in this environment, the dialogue around merger and acquisition strategy has significantly elevated. So while we are seeing deals being delayed, there is a lot more chatter across the industry about inorganic strategies. Well, that poses an issue with capital planning and stress testing, because now you'll have to integrate the other institutions history and practice areas into your plan. So that's something that cannot be forgotten about. It cannot be siloed. 100% must join the inorganic and organic strategies as you're evaluating a possible scenario to your capital plan.
Mike Ensweiler:Well, this has been very insightful and it's certainly great food for thought for the audience. Thank you so much, Macie and Kevin for joining us today.
Macie Fults:Thank you for having us, Mike.
Kevin Kirksey:Thank you so much for allowing us to participate in the discussion, and we're both really excited to help future proof financial institutions.
Speaker 1:I want to thank you again, Macie and Kevin for taking time to talk through capital planning and stress testing in this current environment. At the end of each episode, I'd like to take a moment and let you know about some of the additional resources we have available. If you'd like to learn more about capital planning, you can check out a recent webinar that Macie led along with Thomas Griswold on how much capital does your institution really need in our webinar archive on our website as always stay safe, stay healthy, and thank you for listening to In Your Best Interest: An ALM First podcast.
ALM First:The content in this podcast is provided for informational purposes and should not be relied upon as recommendations or financial planning advice. We encourage you to seek personalized advice from qualified professionals regarding all investment decisions. Current and future holdings are subject to risk and past performance is no guarantee of future results. Podcasts should not be copied, distributed, published, or reproduced, in whole or in part. Information presented herein is for discussion and illustrative purposes only and is not a recommendation or an offer or solicitation to buy or sell any securities. The views and opinions expressed by the ALM First Financial Advisors’ speakers are their own as of the date of the recording. Any such views are subject to change at any time based upon market or other conditions and ALM First Financial Advisors disclaims any responsibility to update such views. These views should not be relied on as investment advice, and because investment decisions are based on numerous factors, may not be relied on as an indication of trading intent on behalf of any ALM First Financial Advisors product. Neither ALM First Financial Advisors nor the speaker can be held responsible for any direct or incidental loss incurred by applying any of the information offered. ALM First Financial Advisors is an SEC registered investment advisor with a fiduciary duty that requires it to act in the best interests of clients and to place the interests of clients before its own; however, registration as an investment advisor does not imply any level of skill or training.