Bob Jensen's Transcript of a Presentation by Walter Teets

In 2017 my Website was migrated to the clouds and reduced in size.
Hence some links below are broken.
Contact me at rjensen@trinity.edu if you really need to file that is missing.

A Presentation by Walter Teets
as Videotaped by Bob Jensen

Bob Jensen at Trinity University

Caveat: . I am grateful to Professor Teets for allowing me to videotape his inspiring presentation. The quotations from Professor Teets that appear at various points in this document have never been edited by him or modified from a transcript of a presentation that I videotaped at a conference. My videotape was transcribed by my secretary, Debbie Bowling. The transcription was modified by me only when Debbie failed to understand certain terminology. I prefer to minimize changes in the transcription so that what is read remains as close as possible to what the audience listened to at the conference. None of us speak with the formalized vocabulary and grammar used in our writing. Also we cannot edit what we said in the same manner that we can edit what we wrote. Bob Jensen added notes in red text.

I also know the least about financial instruments of anybody up here, so I will try not to take too much of your time. But I want to introduce the session a little bit, and then I'll introduce each of the three here. Let me just indicate just whose on the panel, Walter Teets, Rashad Abdel-Khalik, and Hamid Pourjalali. Each one of them, as I said, I'll introduce you in more detail in a little while.

The title of this session is "Financial Instruments and Derivatives." A --- something that probably ten years ago --- I needed to know a little bit about to sound intelligent when I talk to my finance colleagues. But as an accountant, it probably didn't make much difference to me. Well, in the last ten years things have really changed. Knowing something about financial instruments I think is tremendously important to us, when we hear words like forwards, futures, swaps, options, things like that. They used to be things that didn't enter into the accounting statements. Well, suddenly now they do enter into the accounting statements, and they enter in a very complex way.

One of the things, when I think about these, is I think about something that [Joel Demsky--unsure of spelling] told me 30-years ago when I was in a doctoral program that stuck with me. That is, that the whole accounting system, the transactions based-historical cost system was based on a world of certainty. And it works very well in a world of certainty. And in a world of certainty, financial instruments don't become very complex. All they reflect is a time value of money. And all you have to do is account for the timing of the instruments.

Well, unfortunately we have a world of uncertainty. And uncertainty causes complications. And in a sense what we're doing is taking our historical cost transactions-based model, and taking these uncertainty based measures and trying to fit them into a model where they just don’t fit. It's like taking a round peg and trying to drive it into a square hole. I don't think we'll ever get, as long as we stay with the basic historical cost transactions-based model, we'll never get complete agreement on how financial instruments should be brought into this.

So what we have is kind of a makeshift model that brings fair values into a model in which they don't quite fit. I won't try to explain how they try to get them to fit; I'll leave that to the rest of the panel members. But they --- the financial instruments that we've got are a whole variety, very complex. There's an article --- a partner from Arthur Andersen a couple of years ago I think I saw this, had like nine or ten pages just listing names of various financial instruments. And I think there were several hundred names of different financial instruments. But I think if you --- when you look at all of them, they are composed of some certain basic building blocks. Bob [Sworegan--unsure of spelling] in an article said this was kind of like Lego's. You take a bunch of basic building blocks and then you put them together in various ways and you can get just thousands and thousands of different instruments.

But the basic underlying economic principles are fairly simple and there's a fairly small number of them and getting an understanding of those economic principles is really the key to understanding the accounting for financial instruments.

So today we are going to talk a little bit about financial instruments, quite a bit about FASB 133 and accounting for those, and some potential problems that don't seem to be taken care of by 133. The first speaker, Walter Teets, will talk a little bit broader about financial instruments, especially from the SECs perspective. Hamid will then take over and talk in detail about FASB 133. Rashad will talk especially about hedging and some potential problems that come up with the allocation of 133. So that's how we'll proceed. And what I will do is, as each speaker comes up, introduce him with a little bit more detail. So let me start with Walter.

Walter's a member of the faculty at Gonzaga University in Spokane, Washington. He has a Ph.D. from The University of Chicago. He got --- his research areas are generally the impact of accounting information on stock prices and that relationship. But over the last year, he got heavily into financial instruments when he was The Academic Accounting Fellow at the SEC. Just a little plug, plug. The people that I've known that have had that have really gotten a great benefit out of that time at the SEC. So those from the U.S. if its --- if you're interested in this kind of accounting, it's a great way to get some real world experience.

Walter spent the majority of his time working on two different things. One is, the SEC rule regarding disclosure about market risks inherent in derivatives and in other financial instruments. And then more specifically second, in response to and reacting to the various issues involved with FASB 133. So he spent a whole year dealing with financial instruments, and he's going to share some of the insights that he got during that year with us --- Walter.

Thank you, Gary. As Gary said, I'm going to be talking about financial instruments in general and particularly looking at the SECs market risk disclosure requirements. These disclosure requirements in some ways pre-dated the financial accounting standards' boards requirements, but only those really that are implemented in SFAS 133. Because really what most of the requirements the SEC market risks rule include were suggested or recommended disclosures under SFAS 119 which came out some years ago.

We've got a number of things that we can talk about on these instruments, and I'm just going to hit the high points of --- first of all we're talking about market risks. And market risks, the definition of market risks is it's a risk of loss that arises from adverse changes in the market prices. Now these are --- these are not firm, specific things. These are --- I'd guess you'd say economy-like measures that a company doesn't have very much control over, yet they clearly effect their business performance. And we really focus on four types of risks. We're talking about interest rate risks, foreign currency exchange rate risks, and clearly that's the one that at an international accounting conference may be one of the most interest. We're also talking about commodity price risks and equity price risks.

Scope of the SEC's Financial Reporting Release No. 4-08(n) of Regulation S-X

The scope of the SEC market risks rule (primarily SEC Financial Reporting Release No. 4-08(n) of Regulation S-X )includes all derivative financial instruments, which is going to be a major part of SFAS 133 that Hamid and Rashad will talk about; but it also includes other financial instruments. And I'm going to try and focus at least some of my remarks on research issues. Certainly the fact that the SECs rule covers derivatives financial instruments, as well as other financial instruments is an important statement in terms of research. Because one of the problems with looking at derivative financial instruments is that underlying positions may not be captured very well. At least in the interest sense of the instruments, due to the way the market risk rule of the SEC is structured, you're going to have both derivatives and the underlying positions covered.

That's not going to be true of many of the other risks which can cause some pretty serious problems with research perspective. The rule also covers certain derivative commodity instruments, basically those instruments that can be cash settled. Also, the SEC rule recommends that the people include information about underlying positions. That is, if you have commodity derivatives and you're using them in a hedging fashion, then including the underlying commodity positions would certainly be useful to the user. Experience with the first year of filing suggests that most people are not including underlying positions, but at least it is a possibility. There are several pieces to the SECs rule the first is --- that well that all disclose of very quickly is the qualitative information on market risks exposures; and many of the required disclosures under the SEC really translates fairly directly into disclosures under SFAS 133. So you'll probably get a little bit of reinforcement from this later. But first of all, registrants are required to identify a primary market risk exposure. They're to discuss the objectives, the general strategy and any instruments that might be used to manage these market risks disclosures. And they're supposed to discuss what significant changes that may be in the exposures.

SEC registrants, and I do say registrants somce this is not a requirement for all companies, must comply with SEC Financial Reporting Release No. 4-08(n). Thge FASB's requirements cover more companies. SEC rules only apply to SEC registrants. However SEC registrants make up by far the majority of the economic activity in the U.S. Not necessarily the largest number of companies, but certainly in terms of activity. The companies are supposed to present information about market risk exposure in two major categories, trading and [under trading--not sure]. That translates roughly into some categories that the FASB statement requires disclosures on. And registrants are supposed to present information for interest rate sensitive instruments, exchange rate sensitive instruments, commodity instruments, and then others. That other is equity as far as I know. I never heard of anything else. Maybe the other is used just in case someone comes up with something really strange in the future.

I think from a research perspective, the primary things that will be of interest to researchers are the quantitative disclosures that are required in SEC filings. And the quantitative disclosures can be provided in three different methods. And one of the things that I think is a broad research question that we'll come back to a little bit later is, are some of these methods more effective than others? The first disclosure method that registrants can choose is a tabular disclosure method. And under the tabular disclosure method, the registrants should present the fair values of any financial instruments that they have, as well as contract terms that are sufficient to be termed what cash flows would be. And this is really so that the user might be able to do an analysis similar to the two other things that I'll talk about in just a few minutes. For those of you who work in financial industries, the tabular disclosure method is very similar to the gap tables that are already provided by banks. But there's an expansion beyond what the financial institutions have been providing in that there should be information also about currency derivatives, commodity positions and derivatives, and equity positions. And as I said, the objective is to allow the user to be able to come up with what is their [--- outlook--unsure] sensitivity analysis or value at risk analysis which are the two other choices. And that clearly brings up a question in terms of research. Are people able to take this tabular information and do a meaningful analysis on their own?

A second method that companies can use is to use a sensitivity analysis. The sensitivity analysis in essence says if prices change by a certain percentage, what will be the impact on one of three possible measures, either fair values, earnings, or cash flows. One thing that was built into the rule and it was intentionally built in, but causes some unknowns; there's a lot of flexibility built into the rule. There are three disclosure methods that you can use, we've looked at tabular very briefly, sensitivity analysis right now, and in a minute we'll look at value at risk. But not only can you use those three methods, you can also disclose values at risk or sensitivity analysis measures in terms of fair values, earnings or cash flows. And so there's just a tremendous amount of flexibility that is partly in there I think for political reasons, but causes potential problems for users of this disclosure information. And might be something that could be pursued from a research perspective.

The sensitivity analysis in essence is defined as providing something about changes in fair values, earnings or cash flows from rate or price changes that are A) reasonably possible; but what is reasonably possible mean? When I was at the SEC I asked within the office what is reasonably possible mean? And the people I asked are clearly the best technical minds that will be in the profession --- the practitioner profession in the future. And I got numbers anywhere from 51% up to greater than 90%. Now just remember that range when we'll come back to that a little bit later. The change that can occur in the near term, well near term, is defined as within the year. It has to be at least 10%, unless you can justify less than a 10% change. For instance, if you have very stable currency rates, for instance between the Hong Kong gold the U.S. Dollar, you wouldn't have to have a 10% change although earlier today we had someone suggest that you may see big changes in that relationship in the near future. One thing to emphasize is that this is not a stress test. If you've been observing the long term capital management problems, there's a question --- well first of all they didn't have to --- they wouldn't have had to disclose any of this information anyway because they're not subject to these disclosure rules. But even if they had provided some of these numbers, particularly the value of which we'll look at in a minute, wouldn't there have been any notion that there was a major problem developing on long term capital management? And I think the answer is not necessarily, because what happened at capital --- long-term capital management is that in essence very unexpected events occurred, and it hurt for a fairly long time period. And what the sensitivity analysis does is as it says, reasonably possible near term changes in any of these rates.

Well what we say over the last three months, maybe four for long-term capital management and in fact over the last fifteen months with the Asian crisis is unreasonably possible. Or things that just nobody expected that would not have been captured by a lot of the models that have been used in these disclosures methodologies.

The third method that is allowed is indeed value at risk. There's been a lot in print about value at risk and because I can never get the words right unless I can read them, I'll just read what it is. "Value at risk presents the potential loss in fair values, earnings or cash flows." And normally it's the potential loss in fair values that most people think about that with the SEC rule; it does allow fair values, earnings or cash flows from adverse market movements. Now you --- when you use a value at risk, you present the probability of the adverse right rate or price change. And you're looking generally at a 95% confidence, or maybe 99%. And so this is much more specific than the sensitivity analysis. Sensitivity analysis, remember we said anywhere from 51 to greater than 90% changes, here with value at risk we very clearly specify the confidence interval that we're looking at, where the probability that we're looking at.

Time period can be what value change could the company experience over; it could be one day, it could be one month, it could be a quarter, it could be a year if the company wanted to provide that information. And it's entirely of the election of the registrant.

Walter. No, no, what that says is 95 percent of the time we will not observe a loss greater than [unsure], ok? So, out of a 100 days, we would expect to see a loss greater than that only 5% of the time.

That's very different from the sensitivity analysis what we're saying we could see price changes that would lead to this loss, but it's very unspecific as to exactly what we mean by reasonably possible. And I think that that's again a possible research invention because one of the things that I did at the SEC was to talk to a lot of market participants, registrants, financial analysts and so forth. And it's very clear that financial analysts for example think that sensitivity analysis is stress tested. What happens if something really unexpected happens? That's not at all what it is. One thing that might be of interest from a research perspective is to do a study of financial analysts, see what --- see how they're using sensitivity analyses because they have indicated quite strongly that that is their purpose. They would rather see sensitivity analysis and not value at risk. As researchers, we are used to statistical statements of probability assessments. While I say used to, I'm not sure we do it very often in our personal lives, but at least in our research lives we think about those issues quite a bit.

There's a sense that people don't like value at risk, because there is a question about whether it really does provide a probability assessment. The sensitivity analysis looks like point estimate and discloses nothing about variance. So people say well, boy this is a number that I know how to use. I suspect the only thing that happens --- well the real truth is that people don't know what sensitivity analyses is and therefore probably use it incorrectly. And I think it would be interesting for a research perspective maybe for [unsure] research and stuff to see how financial analysis are thinking about this and using that measure that is prepared. For a research perspective, there's another interesting dimension on this rule, I'll move into it now. That is, it was phased in, in two pieces. Starting last year essentially for companies with fiscal years ending after June 15, 1997, and during large market capitalization's they've already had provided this information. The second tier for all other SEC registrants have just started to have to provide this information. And so again from a research perspective, there is some data available. I don't know how much Hamid and Rashad will talk about research on SFAS 133, but a clear disadvantage to researching SFAS 133 at this point is that there's not data in, because it's not required. At least the SEC, which is certainly a compliment to SFAS 133, has data that has started to become available. It's available through the SEC [Edger--unsure of spelling] Website if you're interested. And after this first year, which is already transpired for large companies and is starting now for the smaller companies, people have to present contextual information. What contextual information means is that if you use the value at risk method, you have to tell what was behind value at risk during the previous period. The low value at risk or some distributional information about why you're at risk to put the end of the year value at risks number into perspective.

I've already talked a little bit about the effective date issue, so I won't say anything more about that. The SEC has done a review of the first year of market risk filings; at least part of that review should be up on SEC Website. I don't know whether it is yet, but it will be coming up on the Website. And I should comment that the SEC is very interested in research relating to the market risk disclosure rule. They are interested primarily because Congress is interested. And actually as we were preparing to review April, May, June, July, Congress wasn't very interested. Because Congress has probably --- much of the rest of the population has a very short memory. Back in '94 and '95 this was really a hot issue when there was huge losses than over '95, '96, and '97, there wasn't much happening in terms of derivatives. I think that with what happened with Walter [unsure of next name] Management in the last two months, maybe Congress is a little bit more interested in what's happened to derivatives. Although it is an election year and by the time they get back from elections, they may have forgotten about the derivative issues. But there is a lot of interest in the SEC level in getting research on both their disclosures and on derivatives in general, because they as other government agencies have to --- they have to justify the rules that they pass. And there will be a major review that will occur at the end of three years of the market risks rules existence.

That sounds like a long ways away, but in terms of research perspectives, it's just around the corner. By the time you get data collected and generate working papers out, we're almost too late for SEC needs. So I encourage you if you have interest in financial derivatives or financial instruments research, to consider making use of the disclosures provided pursuant to the SECs market risk rule. I think it will give very useful information.

Things to think about as you're developing research questions, is why were they developed in the first place? Certainly one reason was to reduce information [unsure]. And to provide information to market participants with hopefully the end result of lowering average cost of capital. If you think about risk capital having several dimensions, part of it is for known risks. But there is probably an element of cost of capital that relates to simply the fact that there is an unknown element of risk out there. Now one of the things that we think should happen as companies disclose more information about their market risk positions, is that cost of capital for some companies will indeed go up because we will find --- holy cow --- they really are risky companies. And we didn't know that. But other companies should go down because you find that either they're using these instruments in a correct manner for hedging activities, something like that. They're using responsibility and one would hope that overall the economy wide cost of capital would decrease just a little bit. Certainly there's some research in the voluntary disclosure area that suggests as people disclose costs of capital do go down. This is a great disclosure study possibility. There's a big new disclosure that's being made there. Certainly there's the hope that people will be able to improve their investment decisions and finally that managers private goals could be aligned with investor's goals. And so any of these three let's say purposes of issuing the rules in the first place would be potential avenues for research.

Some specific issues for companies that use tabular disclosures, are these disclosures sufficient to allow investors to do their own sensitivity analyses? Basically to determine the full distribution, the full implications of the financial instruments that companies are using. One of the things that I did at the SEC was to try and take the tabular disclosures of some companies and do a quick sensitivity analysis of my own. And I said, you know if you're going to do this and you want to it for a lot of companies which assumably --- or presumably financial analysts would want to do, you're going to have to automate it some --- in some fashion. And so I built a little spreadsheet making use of [unsure] rate return functions and so forth that would allow me to do my own sensitivity analysis and found out actually that the first two companies that I tried it on, the numbers that they provided were impossible. And we actually sent [continent--unsure] letters to those companies asking what's going on here. And they sent back with, whoops, there were little errors there. And so one of the issues in research prospective is what is the quality of the disclosures in the first place? Is it providing information that is useful to anyone, let alone less sophisticated investors? That relates partly to the third check mark there that even with training, these are phenomenally complex issues, complex instruments and it's very difficult for someone who's A) not an expert and someone who B), has only fairly cursory information to get a very cold picture of what's going on.

Final research question, is this disclosure, does it pass the cost benefit test? You can bet that the SEC is interested in that one because a lot of the registrants are saying, this is ridiculous. We just shouldn't have to comply with this, especially because investors aren't going to really learn anything of use from it.

Sensitivity analysis, disclosure issues --- sensitivity analysis does some things very well. But as I said it kind of gives a [point--unsure] estimate idea, it doesn't a probability assessment idea. Certainly the sensitivity disclosures do not allow a determination of full distribution. That is it says what may happen in a loss situation, but it doesn't say what will happen in gain situations. Additional requirements of the disclosure rules say that if you got highly levered instruments or something like that, you should say that you've shown something for a 10% change, but if it was a 20% change it wouldn't just be twice as large. It might be ten times as large. It's not clear how much of that would be accessible from the sensitivity analysis measures. Clearly since companies are able to choose reasonably possible price changes, some companies are going to choose 10%, some are going to choose 12%, some are going to choose 5%. Is there really comparability across companies? That's a big issue.

Value at risk is primarily a downside measure. It doesn't say anything about what would be the upside potential of these financial instruments. I think Pourjalali is going to talk hedging and I haven't heard his comments, but certainly one function about hedging is there is upside potential as well as downside potential. Value at risk doesn't capture that. Is that a problem? Would investors like information about upside risks? Value at risk is comparable across firms, even that is not quite true because some firms may choose 99%, some may choose 95%, some could choose 90%. Some are going to use one-day windows; some are going to use one month, one quarter. There are a lot of issues there about usefulness of the information. Value at risk does not permit the determination of the full distribution; again it just shows the lower side risk. But maybe in many situations the downside potential is fairly well mirrored on the upside potential as well. So do we have valuable information there?

I think on this light, particularly the second point, will investors over-react to disclosure of potential losses. When you talk about 95% probability of loss, you're talking about a huge amount of loss. And will investors correctly perceive that this is a very unlikely loss, not impossible by any means. But unlikely will investors correctly use that information? I know that some years ago there were --- there was work done particularly with student subjects about do they correctly use statistics, particularly conditional statistics? The answer uniformly was people were just terrible at using statistics. How will people evaluate these values at risk numbers?

Walter. Certainly, and really that was one of the earlier comments that all of --- well, both the sensitivity analysis and value at risk uses in general recent history as the benchmark. And one of the big problems with the Asian crises and with long-term capital management's policy is that the recent pass suddenly was not representative what the firm's [received--unclear].

I indicated that companies were --- or companies are encouraged to provide additional disclosures about underlying positions, but they don't have to. They have a lot of flexibility and so there's a question about whether the flexibility that has provided, has been provided by the rule will in essence make the --- let's say the data --- not the information, the data that's provided under the disclosure uninformative rather than informative. Companies can change disclosure methods in different years. They do have to provide some comparable information from previous years, but again that may cause some problems. Now that's not something that's going to be a researchable question at least for a couple of years, because we won't have any data on it. And then the last question related partly to the [unsure] made, what happens when the environment changes? But we know what happens when positions change quickly, because it's clearly important that financial statements 10Ks, 10Qs don't come out the day after the end of the quarter. They come out sometimes 90 days later for the 10Ks. It's very unlikely a company has the same position at that point. How does that effect the usefulness of these disclosures? All of these I think are researchable issues.

One question is whether risk management activities of companies actually benefit shareholders? The idea of diversification and undiversifiable risk suggests that perhaps the companies should be undertaking a lot of these activities. But the investors should be choosing portfolios on that basis. But clearly managers are undertaking these activities; probably some incentive compensation ideas that might come out of this area. And in fact we've got down at the bottom there might be research that will look at whether you can structure contracts between management and essentially the company that bring into alignment shareholder and managerial interests. And basically change the way firms manage their risks, so you can do research on this issue from a number of different perspectives.

In terms of the SECs interest, particularly and I'm no longer associated with the SEC, but I think that one of the reasons that I was interested in doing fellowship at the SEC in the first place is that I'm frustrated by I won't say a total lack of impact. But the limited impact that academic research has on accounting practice and the SEC is certainly interested in academic research. And I think that the SEC would be interested in research using any of these methodologies to address some of the questions related to market risk rule. If you're a theoretician, there is an interesting paper by Frank O. Wong that does some nice develop as to what types of information about derivatives would be useful. We're in the very early stages of these disclosure rules both for the SEC market risk rule and for SFAS 133 it may be that experiments, experimental markets, experimental economics research could provide some insight into these issues before we have the data. Simulation, I'm very interested in computer techniques of all kinds and it's possible that simulation could be used here where we don't have [a hint--unsure] the data in the first place. And then in [unsure] research will always be at least possibly beneficial but we can't [do it--unsure] at this point because we just don't have the data yet. I'll turn it over to the next person.

Qusestion/Comment. Excuse me, could you possibly tell us where we could get copies of this thing?

Walter. Yes, if you would like copies of the slides get my business card from me afterwards and e-mail me. If you just give me your business card, I will guarantee you it will be lost. It won't be lost long-term, but I won't be able to find it when I get home. So I'll give you my business card with e-mail address on it then I'll make sure you'll get the slides