Podcast Transcript

Speakers

Richard Reed, Vice President, Defined Contribution Practice Director
Jarred Wilson, Vice President and Consulting Actuary
Robert Krebner, Consultant, Princeton

Podcast transcript

Narrator: Governance, operations, investments, compliance. These are the four pillars of risk that define contribution plan sponsors need to stay up on. Every quarter, we're going to be giving you the scoop, the skinny, the lowdown on all the latest developments in the field. Everything you need to know to stay current and informed. We're going to be talking regulatory issues, best practices, investment strategies, all of it about what it all means and more important, what it means to you.

So put on your swimsuit, we're going to be doing some deep dives. Welcome to the Retirement Plans Insider from Segal.

Rick Reed: Hello, my name's Rick Reed and I'm the Defined Contribution Practice Director at Segal. Today I am joined by my colleague Robert Krebner, who is a consultant in our retirement practice. He's joined me in interviewing an old friend to our podcast, Jarred Wilson, who we've talked to in the past. Reason we're talking with Jarred today is our topic is around non-discrimination testing. This is a very important topic at this time of year and one that we think you listeners would find very beneficial.

We will be talking about the required testing and how it works, and really more the common pitfalls and solutions around dealing with non-discrimination testing. So Robert, I'd like for you to introduce yourself and then we'll turn it over to Jarred to say hello. 

Robert Krebner: Hi Rick and Jarred. Thanks for having me. As you mentioned, my name is Robert Krebner. I work in our New York office. I'm a consultant in our retirement practice, and I'm excited to be speaking with all of you today and to talk about this exciting topic. So back to you, Rick.

Rick Reed: So as I mentioned, the reason we have Jarred here with us today is Jarred is our non-discrimination testing expert here at Segal. He has done a great deal of non-discrimination testing with very, very complex situations. So Jarred, why don't you take a few minutes and talk a little bit about your background and how you address non-discrimination testing.

Jarred Wilson: Sure, and thanks guys for having me on this podcast. It's nice to be on the other side of the table for this one. So in my career as a retirement actuary, which I've been doing for 25 years now. Non-discrimination testing, it's one of those areas that really can be a thorn in the side of plan sponsors. It is not something that's ever really a primary focus of the job at hand for plan sponsors. There's typically other issues around managing the plan that are more top of mind, right?

But non-discrimination testing is a requirement for all qualified retirement plans, and by qualified we mean any retirement plan that is subject to the requirements of ERISA, the Employee Retirement Income Security Act, and the favorable tax treatment that comes with sponsoring those plans in order to maintain that favorable tax treatment. You have to satisfy these non-discrimination testing requirements. So what does that mean? Essentially, we are testing to make sure that these plans do not discriminate in favor of highly compensated employees. And we'll talk about what that term actually means.

Rick Reed: Jarred?

Jarred Wilson: What's up?

Rick Reed: Could I interrupt you for a moment? When you talk about plans, could you talk about what types of qualified retirement plans? Are we talking about just 401k plans? Are we talking about defined benefit plans? What kind of plans are we talking about?

Jarred Wilson: So we're talking about both defined benefit and defined contribution retirement plans. And so what we mean by that, a defined contribution plan, the most common form is your 401k plan, right? Where employees will defer their own money into the plan, typically receive some sort of matching contribution from their employer, and perhaps also receive an additional employer contribution that's not tied to employee deferrals that may be a profit share and contribution or for some sort of other non-elective contribution.

And when we say defined benefit plans, those are your more traditional pension plans where a company will provide benefits to employees that they will earn over their career, typically equal to some percentage of the salary they earn during their lifetime. So for example, a common one might be 2% of pay for all years of service. So if you work for 20 years, you're going to get a pension benefit equal to 40% of your final salary paid at retirement.

Rick Reed: Thank you. But today's focus is going to be primarily around 401k plans and 403B plans.

Jarred Wilson: Well, we're going to touch on both because while I think from a timing perspective, what makes this particular topic timely right now is that a lot of companies are doing their testing on their 401k plans because those generally need to be completed by March 15th at the latest. And by completed we mean not only completing the test, but any potential actions that arise from doing that test also need to be completed by March 15th. So that's the primary focus of today's discussion.

However, we are going to touch on defined benefit plans as well because there is some interplay there when we start going deeper into plans that sponsor or companies that sponsor multiple plans. And in many cases, that includes both a defined benefit and a defined contribution plan.

Rick Reed: Great. So if you maybe move a little deeper and let's go into what kind of tests are involved and how does it work. Very high level.

Jarred Wilson: Certainly we're going to keep today's discussion at a high level. For the listeners out there who have had to deal with non-discrimination testing, we know that these things can get pretty deep into the weeds with all of the rules and regulations that are required. But at a high level, I mentioned at the outset that you're comparing highly compensated employees to non-highly compensated employees. And we need to make sure that the plan is not designed to favor the highly compensated employees.

And when we say highly compensated, that's a very specific term that's defined by the IRS based on the amount of compensation that employees earn. The salary threshold that defines whether or not you are highly compensated, it changes each year. It goes up with inflation for testing that is being completed right now, which would typically be a test for the 2023 plan year. You're actually going to look at the year prior, in this case 2022, to see what the salary level is and whether or not somebody is defined as highly compensated.

And the applicable threshold for tests that are being done right now is a salary of $135,000. But again, that amount changes each year and it's going to go up in future years as well. In particular with the high inflation environment that we've seen in the last couple of years. That threshold for determining whether or not you're highly compensated will also increase.

Rick Reed: So in the sense you're kind of drawing a line in your population saying everybody over $135,000 is considered highly compensated, everybody else is considered non-highly. And you're going to be comparing those two groups.

Jarred Wilson: Exactly. We compare those two groups. And one other thing I do want to add, for the most part, when we talk about who's highly compensated, we look at that salary threshold. There are a couple of other components that could define somebody as being highly compensated. One important one is ownership. So depending on the type of organization that we're working with, if somebody is more than a 5% owner, then regardless of what their actual salary is on a W2 basis, if they're a 5% owner, they would also be considered highly compensated.

Rick Reed: I see.

Jarred Wilson: But to your point, we're drawing that line and saying, "Okay, we have this group over here and we have this other group over here, and we have to compare between those two groups." How many of them are participating in these plans? To what degree are they participating in, particularly with 401k plans? How much are the highly compensated employees deferring of their own money versus the non-highlys? And then for the employer provided benefits, is the formula set up such that your highly compensated employees are getting more of a benefit than your non-highlys? And if that's so, that's where some of these testing issues can come into play.

Rick Reed: So what are these tests? Because a lot of times you hear about the comparison, I'll just say we hear about ADP and ACP testing. Is that all that's done or are there others?

Jarred Wilson: So those are the two most common for sure for 401k clients. Those are the tests that, we had mentioned earlier, need to be completed, and any corrective actions if necessary, would also have to be completed by March 15th. Your ADP test, which stands for actual deferral percentage, that tests the employee contributions that are made to the plan. So contributions that employees are taking out their own paycheck and deferring typically on a pre-tax basis, although there's also a special kind of contribution referred to as a Roth contribution, which is made after tax.

But then your earnings on that money grow tax-free into the future. So both of those types are included in this ADP test. And then the ACP test for actual contribution percentage, that's where you test your matching contribution, the money that you're receiving from your employer as a direct match to your employee deferrals. So we have to test both of those amounts. And again, keeping us at a high level, not getting into the nuances of how the calculation is done, is very specific formulas that we look at to determine whether or not the plan is deemed to be discriminatory.

But we're taking that calculation of the employee and employer money for the highly compensated comparing to the non-highly compensated and seeing if we have a discriminatory issue. But in addition to those two tests, while those are the most common, there are other tests that are required for 401k plans and other plans as well that we mentioned at the outset. A profit sharing plan as an example, or a defined benefit pension plan. Those tests are centered around such things as coverage, which means how many employees are actually participating in the plan and how many are not.

Rick Reed: Is it participating or benefiting?

Jarred Wilson: Well, it's benefiting, and that's actually a good question or a good point that you raised because for example, in a 401k plan, you may be a participant in the plan, meaning you're eligible to defer some of your own money, but you may choose not to. And so for the coverage test, this test that looks purely at counts, are you benefiting or not? For that test, you're actually considered benefiting even though you're not choosing to actually defer any of your own money.

But then for some of the other plans, like a defined benefit pension plan, you actually need to be receiving a benefit from the plan. You could be considered a participant in the plan because you previously were earning a benefit, but if the plan design has now changed and you're no longer earning any additional benefits under that plan, you're not considered benefiting anymore. And that could hurt this coverage calculation, this ratio calculation that we do.

Rick Reed: That's helpful. Robert, do you have any questions for Jarred at this point?

Robert Krebner: No, no, let's keep going. It's all good.

Rick Reed: So that was a really good explanation in high level of a complex situation that comes about every year. So let's move on to the theme of this discussion. Let's talk about some of the common pitfalls that plan sponsors encounter and how to deal with them.

Jarred Wilson: So I think at the outset, when a plan is originally set up by an employer, I think most often that's not where the issues arise. They don't arise at the beginning, right? Where a plan is generally set up to be somewhat equitable across the population. But what we see is over time things happen to the makeup of the company and the plan itself, maybe multiple plans come into the mix, and this is where issues can start to arise. So one example of that is during a merger and acquisition.

So if a company has been operating on its own, let's say for several years, but then at some point in time it looks to acquire another organization or maybe it becomes acquired by another organization and now suddenly you have separate populations but still technically falling under the same organization, they're part of the same controlled group is how we refer to that. And once that happens, now you have employees receiving different benefits, whereas maybe company A sponsored a defined benefit pension plan and that was it. They didn't have a 401k plan, they just acquire company B.

Company B has a 401k plan, but they don't have a defined benefit pension plan. That's a very common situation where half your population's getting a pension benefit, the other half is getting a 401k matching contribution, and the level of those two benefits can be vastly different. And you have to test those populations together as if it's all part of one group. And so we see that situation occurring a lot. And typically when an organization is considering a merger acquisition, the last thing they're thinking about is how is this going to impact my non-discrimination testing?

There's a significantly larger number of business reasons that are being looked at when a company is considering M&A activity and non-discrimination testing is way at the bottom of that list. So it gets overlooked and then next thing you know, a year or two down the road, suddenly the auditors come in, look at the plans and start raising questions that if not done properly, nobody at either organization has thought of. And that's where we come into to help things out.

Rick Reed: So let's talk about some other common pitfalls mean. You mentioned that one that's maybe unique to some organizations. What are some more common pitfalls? Just for example, what happens if a plan fails the non-discrimination test?

Jarred Wilson:  Another common one, and this one is very common with the 401k plan in the ADP test. So recall, the ADP test is measuring employee money, how much employees are putting away out of their own paycheck into their 401k plan. And not surprisingly, your highly paid individuals more often than not, will contribute more money to the 401k plan than the lower paid individuals. And there is a threshold there that these tests do allow your highly compensated employees or they allow employees in general to put away varying amounts of their own dollars.

And for those that are highly compensated, it doesn't have to be completely equal to the non-highly, there is a threshold built into the test that allows them to put a little bit more away, but there's a limit to that. And it's very common that when we get to the end of the year and we compare the amount of contributions made by highly compensated employees to those made by non-highly compensated employees, that we fail. That the HCE... HCE is just shorthand speak for highly compensated employees.

The HCE have put away much more money than the non-highly compensated employees. And now what we find ourselves in is a situation where we have to refund some of that money back to the highly compensated employees. They put too much away. The non-highlys either didn't put enough away or maybe didn't contribute at all because they're living paycheck to paycheck, whatever the situation may be. So now our highly compensated employees are limited in what they actually can defer. And if they deferred more than that, then refunds are required by March 15th of the following year.

So for example, any organization that's doing their 2023 testing right now, it's very common that sometime during late January, early February, the tests are being done. They realize, "Okay, we have to refund 20 of our highly compensated employees some amount, a thousand dollars each or some percentage of pay each." And there's a formula for determining what that amount is. So not only do you have to refund that money, you have to refund all the earnings on that money.

And this can be a real headache for employees who are also trying to do their personal income taxes right now and are basing that income tax calculation with their accountant. They're including their 401k deferrals in their own taxes, and now they're getting a refund of some of that money, and it just creates headaches all around. And that's a very common scenario that a lot of our clients find themselves in.

Rick Reed: So if a plan fails, any of the non-discrimination testing, it's just not... You can't just say, "Okay, well we fail and move on." Some sort of action has to be taken, right?

Jarred Wilson: Correct. I mean, we would say nobody ever truly fails a non-discrimination test because if the initial determination indicates a failure, well, that just really means now we have to take additional steps, look at additional tests, whatever is needed to get to a pass. And sometimes that would require a refund, as we just mentioned, for the ADP test. For most of the other situations, if there is an initial failure in the test, then what happens is you can't take away money that's already been earned by any employee.

So let's say we're looking at a test for a profit sharing contribution, and we're testing the contributions that have been provided by the employer to all the employees. So in this situation, if the test tells us that at the moment we're failing. We can't refund those contributions to the highly compensated employees like we just discussed when we're talking about their own deferrals. In this situation, the only way to correct that test is to give more contributions to the non-highly compensated employees.

And that's also a fairly common occurrence that we'll see where the plan design may be set up. But say there's a plan design that by its very structure gives larger contributions to longer serviced employees. There's service bans that are built into the contribution levels. That's a great design. You're rewarding to people who have been there for a long time, but oftentimes the people who have been there the longest are also the highest paid.

And you may be in a situation whereby the plan's designed through no ill intent, it just so happens now that all of the higher paid individuals are getting larger contributions that may result in an initial failure. So now you got to go back to some of the lower paid folks and bump them up at least until you get to a point where the test result is a pass.

Rick Reed: So you have to do things to set things right. You have to pass. What I'm hearing is you either have to bring highly compensated people down on average by giving them refunds or bringing the non-highly compensated people up, which could be costly.

Jarred Wilson: And that idea of bringing the highly compensated employees back down again, that's really when we're just talking about their own money, right? That we're going to do a refund of the employee deferrals. A lot of these tests, if we're testing a defined benefit pension plan or a profit sharing plan, the key thing there is if somebody's earned that employer provided benefit, you can't take it away. So there the only course of action is really giving additional benefits to the non-highlys.

Rick Reed: And I guess my comments are more focused from a 401k perspective. So let's also talk about... Can you tell us some measures or some proactive measures? We've just talked about some reactive measures if a test fails. Can you talk a little bit about some proactive measures that employers can take action so they can at least either guarantee they pass or at least improve their chances of passing by the end of the year?

Jarred Wilson: For sure. We encourage all plan sponsors who find themselves unexpectedly in this situation to... If you have one year where you have to do some sort of corrective action, we strongly recommend, "Okay, now let's take a look at our entire plan design and see what can we do to change our plan structure to never have to deal with this again." And for plans hopefully, who have never even gotten to that initial failure, we like to have those conversations and say, "Hey, here's some options you should consider."

Because down the road, you might find yourself in a position where you may have a potential failure and have unexpected additional benefits that you suddenly need to provide. So the safest option is to design a plan that's considered to be a safe harbor plan. That term safe harbor, if you meet the requirements of a safe harbor plan, that by definition means you are not subject to the non-discrimination tests. And again, there's very specific formulas that have to be adhered to in order to qualify as a safe harbor plan in particular with 401k plan designs.

For a defined benefit pension plan or a profit sharing plan, generally your safe harbor, everybody gets the same amount. So if it's a profit sharing contribution, everybody in the company gets 5%, let's say. That's safe harbor, there's nothing to test because every single person gets the same benefit. That's a guaranteed pass. On the 401k side with the employee deferrals and the matching contributions, there's some nuances there to what the formula can be.

But in general, if you have a safe harbor plan design, you don't have to worry about anything that we're talking about today. You're not subject to non-discrimination testing,

Rick Reed: But you still have to collect data and have your record keeper or your consultant, whoever you use, a company still has to do the testing. They may get an automatic pass, but they still have to do it to make sure [inaudible 00:23:28]

Jarred Wilson: Well, for sure, you have to ensure that the plan is being operated in accordance with the terms of the plan document, right? So if you design a plan like safe harbor, but you're actually providing a matching contribution that is not in line with what the plan document says it is, then you're in trouble. So you still always need to keep track of the record keeper's information and making sure that the employee deferrals and the employer matching contributions, et cetera. We haven't mentioned the word vesting yet.

That's another concept where there are requirements around how employees become vested in the money that they earn from their employer. And by that we mean in certain situations, if you get a matching contribution, but then you leave the company a year later, you may forfeit that match. But under the safe harbor rules, there's very specific requirements on when they can no longer take that money back from you, even if you leave after a year or two years, whatever the case may be.

So beyond safe harbor though, because while safe harbor guarantees you a pass of non-discrimination testing, it may not provide the level of retirement benefits that you truly want to give to your people. A plan design should not solely be done just to satisfy testing. Ideally, all of these retirement offerings should be in line with the goals of the company from both an HR and a finance perspective. So you may want to have, I mentioned before, a service-based design that provides more retirement benefits to people who have been with the company longer.

And there's a number of different designs that can be done around age and service and which department you work in. You're allowed to have these different design structures, and you may want to for purposes of how you run your business. But if you do that, then what else can we put into the mix to at least protect you for non-discrimination? On the 401k side, one of the most common ones is auto enrollment. Where the employee's hired, and without any type of formal election on their part, they will automatically be put into the plan and automatically have some level of a predetermined percentage of their paycheck put into the 401k plan.

And not only is it a great way just in general to get people to save money for retirement, which is really why auto enrollment was put into effect. But it also can very much help with your testing because oftentimes a common cause of failure in that ADP test are the non-highly compensated employees that don't put any money away. So this really solves that issue. And then on the flip side of auto enrolling everybody and automatically putting people in at some sort of minimum threshold. On the flip side, we also can put in a cap, a limit on how much money the highly compensated people can put away.

Now with that accomplished, you don't necessarily want to limit what your highly paid people can put away into their own 401k plan. But if you know every year you've been dealing with this issue of having to give them refunds and all the tax implications that come along with that, you can put in a cap that at least eliminates that issue. And now they know, "Okay, I maybe want to put away the full amount that's allowed under these plans." Which I forget the exact limit for 23, but it's around $22-23,000 per year. Well, maybe they can't do that. Maybe put a limit of $15,000, whatever the case may be, to ensure a pass.

Because you know based on your population any time people start putting in more than that, they end up having to get a refund. So you put in that cap, you avoid the testing failure, and then you look for other ways to potentially make up for this cap that you've put in place.

Rick Reed: Thanks. Those are really interesting points. I've seen different clients over my years take different actions, and I guess the good thing is you also don't have to stay down the same road for the rest of the plan in life. You have options, and there's different scenarios that you can choose on, I don't want to say on a year by year basis because you want to be consistent. But as I understand it, there's different alternatives, both reactive and proactive measures. One other thing I want to talk about, and this will bring Robert more into the conversation, is we're talking about comparing all these employees.

So any employer or any organization that has hundreds or even thousands of employees, you're doing the testing based on your data. Let's talk a little bit about the data. I mean, what can we say about data? Is it critical?

Robert Krebner: It's absolutely critical. The thing is, if the data's incorrect when you're doing the testing, you may need to rerun those tests, and that could increase the cost for the plan... It's just further headache for the plan sponsor. It just drags out that process even further. Absolutely. And in an ideal setting, all records are really consolidated into a single administration quality database. And if that's the case, then the plan sponsor can run reports from one system and they have all that data in one place.

But if participants are moving between plans, it could be difficult to capture that. And that's why one database could be really critical and can really help mitigate some risk there of misplacing certain participants. The other thing is, if you have multiple systems and the HRIS isn't configured properly, there might be limitations about how the flow of data is occurring. And that might require someone to manually push a bunch of buttons in order to conduct that testing. And then you're much more prone to those operational risk and human error, things you just wouldn't think would happen until all of a sudden it does.

And Jarred, you mentioned control group, and when that happens, my mind immediately goes to multiple record keepers. It is possible that if you have multiple record keepers, the record keepers might not be aware of other plans in the control group, and NDT, as you mentioned, must take into account every plan while doing the testing. So it's really critical that record keepers are aware of the other plans and take responsibility for those plans, even if they rarely interact with them to make sure that the testing is being done properly.

Jarred Wilson: You've touched on a lot of good points there, Robert. We can only do the test based on the data provided. If that data turns out to be inaccurate, then the test becomes inaccurate or invalid or practically really meaningless. And you raise a lot of good points on the common pitfalls we see based on multiple systems. Mergers and acquisitions, which we've touched on a few times during this call, just so many potential issues can arise when you bring multiple groups together that are most likely in multiple locations, potentially on different parts of the country.

So now you're looking at different systems, you're looking at different benefits being provided, and that data is not centralized. And when we look at the design of a plan, we talk about this concept of harmonization a lot, right? Harmonize your benefits. That it's true for whatever reason over time, you've now have these different benefits structures in place. It can make a lot of sense to try to harmonize those benefits. And the same thing can be said with the systems that are in place as well, and the vendors that you're working with and your record keepers.

There may be a period of time which you just have to have multiple record keepers for whatever reason, because it's just not something that can be focused on. And there's a lot of work that goes with merging all the information from one system into another. But at some point that the harmonization of those systems is critical because the term garbage in, garbage out exists for a reason. That if that data comes in and it's poor and it's inaccurate, the results that you derive from that data ultimately become meaningless.

Robert Krebner: And you mentioned multiple locations. I mean, we see that a lot in the healthcare industry with different hospitals and higher education where you have different campuses. A lot of times those different locations use different systems to administer those plans. So it can become really challenging to maintain that good data and have the communication that you need. And you just need a really strong effort from the entire team to make it all work properly. And they all have to follow the same rules.

They have to correctly monitor against plan limits. They have to monitor for transfers from one location or one plan to another. So there's just a lot of people and databases that have to be speaking to each other properly in order to make this happen. So that brings me more to a comment than a question in that what I'm hearing is, while many record keepers do the non-discrimination testing, it's part of the fully bundled package of services that they offer any of their clients. What I'm hearing is it kind of takes planning.

It takes planning on the organization's part to make sure that the right data is being used, that the design of the plan is being followed, that all the plans are being considered and tested together if need be. So it sounds like it shouldn't be just, "Oh, my record keeper's doing it for me. They have whatever's on their system." It sounds like they really need to take a proactive action to make sure that it's done properly and accurately.

Jarred Wilson: And I was just going to add, the record keeper, they only know what they know. They have what they have. They don't know what they don't know in a lot of situations. And they may be, as you said, as part of their bundled package, they're doing a test based on the data that they are administering. But there oftentimes can be other data out there within the organization, other plans that the record keeper's not even aware of. And therefore, again, the test that they run is on only partial information and ultimately could be meaningless because it's not truly capturing everything that the test require to be captured.

Rick Reed: Well, I know we can go on and on about this topic because there's so many tentacles and scenarios, but our time is somewhat limited. There are some other things that we could talk about, but maybe we'll save for another episode. For example, if a highly compensated employee gets a refund, are there other alternatives such as non-qualified plans? And obviously the answer is yes, but that's too much information. We can have that for another discussion. Are there any last minute comments or thoughts that either of you want to add?

Jarred Wilson: I think I would just close with, to your point, once we go down this rabbit hole of all the nuances involved with non-discrimination testing, we can go on for hours and hours. But really the key takeaway is just that if you work for an organization, if you're working for a company that sponsors any type of retirement program defined contribution or defined benefit. You need to be aware that these testing requirements exist, that they're complex. And that just because your record keeper is giving you a very streamlined version of a test that may or may not actually capture everything that is required.

And if at some point you find yourself in an audit, the IRS will randomly audit plans all the time, you may suddenly discover that there are things missing that you weren't aware of. That's what we're here for. Reach out. We're happy to talk with anybody that needs help in this area to just do an overall assessment of what are your plans, who are your people, who's benefiting? And do we think the structure as a whole is satisfying the requirements? Or do we see areas that potentially might need to be cleaned up to avoid larger headaches down the road?

Rick Reed: One other point I'll make, and then we can wrap this up. I just want to put it out there that we talk about data collection and how important it is. A lot of times with record keepers, and again, nothing against record keepers, they do a great job, is all this data that we're talking about for non-discrimination testing. This data is used for a lot of different things from the plan year perspective. In other words, they use the same data when putting together the Form 5500. So I just want to make that point that the data is critical more than just for testing. It's used in other areas as well.

Robert Krebner: Absolutely. And NDT is just one of those reasons why it's so critical. So this is a topic that should not be overlooked.

Rick Reed: Well, Jarred, Robert, thank you for this time. This was great information. I hope our listeners really got some good information out of it. I'm sure they're walking away with even more questions, but that's the purpose of these discussions, to intrigue and to think things through. So I hope you found it all helpful and hope you all join us for our next podcast, which will be next quarter. And thank you and have a great day.

Jarred Wilson: Thanks, Rick. Thanks, Robert. Talk to you guys soon.

Robert Krebner: Thank you both.