Jump to content

overzealous auditors


Recommended Posts

Is it an industry trend for pension plan auditors to go nuclear and penalize a plan sponsor with all kinds of Notes for seemingly minor data discrepancies such as one participant's date of birth off by a year out of 400 total participants, or a gender being incorrect on 3 participants?

Link to comment
Share on other sites

If the auditors are just putting that BS in the Management Letter you can explain it away as just their attempt to justify the outrageous fees they charge (they have to find something wrong).

If they're putting anything in the footnotes to the financial statements which get filed with the 5500 that's a whole different ballgame and they should be called to task (throw back in their face the wonderful accounting principle of "materiality).

Link to comment
Share on other sites

We've had a couple of clients have auditors go nuclear over things that were very minor and things that were actually correct, but the auditor thought they were wrong.  They seemed to be on a feeding frenzy trying desperately to find something they could say was wrong.  Guess which clients hired a different auditor the next year?

Link to comment
Share on other sites

Just a point of curiosity:

 

If the data about participants is the employer’s data, what distinct source of information (if any) does an independent qualified public accountant compare that data to?

 

Or if the only data is the employer’s data, what methods does an IQPA use to find an incorrect entry in the employer’s records?

Peter Gulia PC

Fiduciary Guidance Counsel

Philadelphia, Pennsylvania

215-732-1552

Peter@FiduciaryGuidanceCounsel.com

Link to comment
Share on other sites

37 minutes ago, Fiduciary Guidance Counsel said:

Just a point of curiosity:

 

If the data about participants is the employer’s data, what distinct source of information (if any) does an independent qualified public accountant compare that data to?

 

Or if the only data is the employer’s data, what methods does an IQPA use to find an incorrect entry in the employer’s records?

The employer could have conflicting data in its files, especially if a lot of it is done by manual input.

The employer could have given its vendors (TPA or Actuary for example) information with a typo, or the vendor got the correct information but a mistake was made when it went into its records.  Now you have plan data and employer data that is different.  

I don't think the IQPA would have an outside source to confirm that the employers data is indeed accurate data. 

 

 

Link to comment
Share on other sites

First, not an auditor and second, not taking the other side - and you may know for sure (or maybe not) that it was one DOB in 400 that was incorrect, and presumably 3 genders among the same count, BUT the auditors had, I'm sure, a much smaller sample size and so this looks more material to them than it does to you or the plan sponsor. If I'm doing QC and I see 3 out 10 problems, I'm going to conclude that there are issues or control concerns with +/-30% of the population rather than think I happened to sample all or a majority of the issues by chance.

Kenneth M. Prell, CEBS, ERPA

Vice President, BPAS Actuarial & Pension Services

kprell@bpas.com

Link to comment
Share on other sites

4 minutes ago, CuseFan said:

If I'm doing QC and I see 3 out 10 problems, I'm going to conclude that there are issues or control concerns with +/-30% of the population rather than think I happened to sample all or a majority of the issues by chance.

Is it really reasonable to conclude that there are control concerns with +/- 30% just because your random sample had that issue?  Wouldn't a reasonable approach be that the discrepancies in the random sample trigger a larger sample?

 

 

Link to comment
Share on other sites

28 minutes ago, RatherBeGolfing said:

Is it really reasonable to conclude that there are control concerns with +/- 30% just because your random sample had that issue?  Wouldn't a reasonable approach be that the discrepancies in the random sample trigger a larger sample?

What about the flip side of that.  Just because your small sample contain few if any errors, would it be reasonable to interpret that there are little if any errors throughout the entire population?

QKA, QPA, CPC, ERPA

Two wrongs don't make a right, but three rights make a left.

Link to comment
Share on other sites

Remember, data gathering for audits tends to be assigned to the newest team members.

Also, Why are you not verifying the data you received for the other 399 participants?  I think you need to prove to your client that you have addressed any and all issues.

Kristina

Link to comment
Share on other sites

As mentioned in the reply from Lou, the issue isn't really the number of data discrepancies, it's the materiality.  For example, it might be possible the male/female identifier is wrong for 50% of your population and the resulting liability is off by 1% (or less).  While sampling is a good start, it does not tell the entire story.  Talk to the actuary!

I'm a retirement actuary. Nothing about my comments is intended or should be construed as investment, tax, legal or accounting advice. Occasionally, but not all the time, it might be reasonable to interpret my comments as actuarial or consulting advice.

Link to comment
Share on other sites

16 hours ago, BG5150 said:

What about the flip side of that.  Just because your small sample contain few if any errors, would it be reasonable to interpret that there are little if any errors throughout the entire population?

It could be reasonable, sure.  If audit sampling is appropriate to the objective, you aren't looking for absolutes, you are looking for a certain degree of assurance.  

 

 

Link to comment
Share on other sites

  5 hours ago, RatherBeGolfing said:

Is it really reasonable to conclude that there are control concerns with +/- 30% just because your random sample had that issue?  Wouldn't a reasonable approach be that the discrepancies in the random sample trigger a larger sample?

Quote

What about the flip side of that.  Just because your small sample contain few if any errors, would it be reasonable to interpret that there are little if any errors throughout the entire population?

RBG has it right.  If you have problems in your small sample, you pick a bigger sample.  I used to be a plan auditor, and that's audit 101.  Now, part of this is that the auditors lack the requisite expertise in auditing 401k plans.  Oh they think they know everything they need to, but the reality is that many many do not.  Now, some are awesome at this work don;t get me wrong.  But a lot only don't realize they haven't got a clue.

Also, auditors send confirmation letters to participants home addresses directly to confirm information.  So that is how they would get independent corroboration that the data is correct.

Austin Powers, CPA, QPA, ERPA

Link to comment
Share on other sites

I'll speak up for both sides and say if it is simply a management letter comment that does not rise to the level of a 5500 disclosure, then that's okay--if they are making a bigger mountain out of a molehill, that's another thing entirely---their new dictates from the audit guidelines require them to be even more detail-oriented---that said what you describe should definitely be more of a molehill than a mountain and should be treated as such as any conscientious auditor--I became a CPA by taking night classes and I've supervised the tax side of several thousand plan audits for nearly 2 decades for a Big 4 and while the world is changing, that's the way it still should be. 

Link to comment
Share on other sites

14 hours ago, austin3515 said:
 

RBG has it right.  If you have problems in your small sample, you pick a bigger sample. 

So, if you have little or no problems in your small sample, you move on?

QKA, QPA, CPC, ERPA

Two wrongs don't make a right, but three rights make a left.

Link to comment
Share on other sites

3 minutes ago, BG5150 said:

So, if you have little or no problems in your small sample, you move on?

at a very basic level, you set sample parameters so that that you can expect your sample to be representative of the population.  If the sample is representative of the population, you would have a reasonable basis for a conclusion.  

 

 

 

Link to comment
Share on other sites

2 hours ago, RatherBeGolfing said:

at a very basic level, you set sample parameters so that that you can expect your sample to be representative of the population.  If the sample is representative of the population, you would have a reasonable basis for a conclusion.  

 

Except if the results are bad.  Then you expand the selection, it seems.

QKA, QPA, CPC, ERPA

Two wrongs don't make a right, but three rights make a left.

Link to comment
Share on other sites

5 minutes ago, BG5150 said:

Except if the results are bad.  Then you expand the selection, it seems.

Of course. You want your audit to tell you where/what the problems are, not just that there are problems. 

lets say that an audit is supposed to make sure a big report is free of errors and misstatements. Some things have so many details that it is either impossible to go line by line, or it would be incredible inefficient and cost prohibitive.  Instead, you use audit sampling to draw a reasonable conclusion.  If you find problems in your sample, you get a bigger sample and continue to dig until you find all problems or are reasonably sure that you have found all errors.  If you don't find problems in your sample, you can draw a reasonable conclusion from the sample.

This all assumes that audit sampling is appropriate for the objective and that the auditor follows proper protocol that includes risk tolerance and all that fun stuff.  For example, I wouldn't just check 10 participants and say they are all good so I can conclude that all participant received their contributions.

 

 

Link to comment
Share on other sites

To follow up with some of the comments:

The circumstance here is indeed two different data sets - the original employer source census files and the actual annual valuation files that are imported into the valuation software.

The discrepancy is based on comparing the two full sets of data which was a reasonable task rather than a more complex verification that the employer's source data itself is indeed correct.

That's why we are feeling it was a nuclear response. Don't know yet the final iqpa conclusions. They have threatened to put in Notes but we are yet to see.

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...