It may be said that the long-term goal of investing is to multiply the eggs in our baskets. Yet too many investors focus on producing more eggs (getting high returns) while paying little attention to the fox (costs) that perpetually robs the hen-house. If you ignore the fox, soon there will be nothing left to produce more eggs.
—Scott West and Mitch Anthony, Story selling for Financial Advisors
John Bogle, founder and former chairman of the Vanguard group, is one of the legendary figures of the investment business. In a keynote speech, “Investing with Simplicity,” given in Philadelphia on October 3, 1998, Bogle made the following statement: “The realistic epitome of investment success is to realize the highest possible portion of the market returns earned in the financial asset class in which you invest—the stock market, the bond market, or the money market—recognizing and accepting that that portion will be less than 100 percent.” One of the critical messages Bogle was trying to convey was that costs matter. In fact, because of their negative impact on returns, they matter a lot.
It is vital for investors to understand that the costs of trading fixed-income securities can be greatly affected by the manner, and market, in which they are traded. Most investors are probably not even aware that the bond market is actually made up of four distinct segments, each with its own unique characteristics. The unique characteristics have important cost implications for investors. There is the primary market and the secondary market. There is also the interdealer market (wholesale) and the retail market (end investor).
Bonds that are purchased at issuance are said to trade in the primary market. The initial offering of a U.S. Treasury debt instrument is sold through an auction. Individual investors can participate in this process without incurring any transaction fee by establishing an account with the Federal Reserve Bank. This is done through what is called the TreasuryDirect program. This program, however, is available only for taxable accounts. Alternatively, Treasury issues can be purchased through banks or brokerages, in which case a modest fee of $25 to $50 is typically charged, regardless of the amount of the purchase.
The sales of the initial offering of corporate and municipal bonds are done differently. They are not sold directly to the public. Instead, they are sold to the public through an investment bank acting as what is known as an underwriter. Often groups of investment banks band together to form what is called a syndicate to buy the bonds. The syndicate then generally acts as part of a selling group that markets the bonds to the public. All bonds are sold at the same price. This is one reason why it is generally preferable for individual investors who do not have access to the interdealer market to buy individual bonds in the primary market— there is no markup or commission taken by the seller as the selling “concession” is paid for by the issuer. Thus individuals can be assured that they are getting the same price as large institutional investors.
Once the initial offering is completed any trading is then done in what is known as the secondary market. We begin by examining the secondary market for U.S. Treasury debt. The secondary market for Treasuries is the most liquid market in the world. In the wholesale (interbank or interdealer) market the typical trade is in blocks of $1 million and trades as large as $100 million are not uncommon. Just as is the case with equities, the prices of out-standing Treasury issues can be found in the Wall Street Journal. Thus prices can be said to be transparent. However, once we move beyond Treasuries, and enter the world of the secondary market for corporate and municipal bonds, the markets become opaque—at least at the retail level.
The Federal Reserve Flow of Funds report has estimated that while individual investors directly hold 40 percent of the market value of all corporate equities, they directly hold just 7 percent of the total market value of debt instruments. With institutions controlling 93 percent of the market, the technology related to information on pricing of fixed-income securities is designed for the institutional market. The technology that allows professional investors (both broker-dealers and fund managers) to follow market pricing in a timely manner is very expensive, and thus generally not available to the individual investor. The expense is so great that even most financial advisors do not have access to the pricing data. The result is that there is a large asymmetry of pricing knowledge between retail and institutional investors. This asymmetry has major implications for the prices that individuals pay.
Whenever there is a large asymmetry of pricing knowledge between retail and institutional buyers it is highly likely that the prices paid for the same product will vary greatly. An example of how an asymmetry of knowledge impacts prices paid is the diamond business. The markups between retail and wholesale prices paid for diamonds are typically very large. If you shop around, you will likely discover that diamonds of essentially the same quality (i.e., cut, clarity, weight, and color) are being sold by different retailers at vastly different prices. For example, if you happen to have a good friend who works in the diamond district in midtown Manhattan the price is likely to be dramatically lower than what it would be at most retailers. In any case, it is unlikely that the buyer will ever know the true wholesale price. Similarly, an asymmetry of knowledge has great implications for the prices that individuals pay when they buy securities. Consider the following.
When an individual buys shares in a company (e.g., Microsoft) whose stock is traded on one of the exchanges, she can buy that stock from any broker. She also can easily find out exactly what the bid and offer prices are for the stock at any given moment. The price at which the purchase will be executed is likely to be very similar, no matter which firm executes the trade on her behalf. The only difference might be the size of the commission. The transparency in pricing protects most investors from being exploited in terms of the price they pay. This is, unfortunately, not the case in the municipal bond market. Nor is it the case with corporate bonds, with the exception of the few corporate bonds that trade on the New York and American stock exchanges (NYSE and AMEX). Contributing to the transparency of prices is the centralized manner in which trading is conducted. The market for municipal bonds is a very decentralized one. It operates through an over-the-counter network of primary and regional dealers. The lack of centralization contributes to the lack of transparency in pricing.
There are other crucial differences between the secondary market for equities and the secondary market for bonds, especially municipal bonds. The first major difference, as we just discussed, is that while individuals directly hold about 40 percent of equities, they directly hold only about 7 percent of bonds.
The second major difference is that while stocks are traded very heavily in the secondary market (over a billion shares are typically traded daily on the NYSE alone) most municipal and corporate bonds are owned by buy-and-hold investors. Thus while the equity market is highly liquid (resulting in low trading costs), the market for municipal and corporate bonds is generally very illiquid. The lack of liquidity results in very high trading costs.
The lack of liquidity in the secondary market creates another difference between stocks and bonds. If you want to buy any particular stock, you can do so simply by calling a securities dealer. The situation is very different for municipal bonds. It is not likely that a dealer will actually be holding in inventory the specific bond you want. They will have to search the market to find someone who owns the bond. And they may not be successful. The more difficult a bond is to find, the greater the price you are likely to pay. Fortunately, this does not present much of a problem on the buy side since bonds (at least investment-grade bonds) of the same credit rating and maturity are highly likely to provide the same return. The problem, however, is that the lack of liquidity and transparency makes it very difficult to compare prices of securities of comparable rating and maturity. The liquidity and transparency issue becomes much more critical if you have to sell a municipal bond prior to maturity—you are likely to incur a significant cost in the form of a price concession (markdown or “haircut”).
A third major difference is that the secondary market for individual stocks generally remains vibrant as long as a company is public. This is not the case with municipal bonds. The longer a bond is outstanding, the less frequently it tends to be traded. Some bonds may not trade for months or even years. Another related problem is that it may be very difficult to obtain the prospectus for a bond trading in the secondary market.
Finally, the sheer number of issues creates a problem in reporting the prices of municipal bonds traded in the secondary market. While there are about eight thousand stocks traded on the three major exchanges (NYSE, AMEX, and NASDAQ), it is estimated that there are about 50,000 issuers with over 1,700,000 individual municipal bond issues outstanding (more than all the public equity issues outstanding around the globe). According to the Municipal Securities Rulemaking Board (MSRB) only a tiny fraction (1,000 out of 1,700,000) of these issues trade on a daily basis. Only 10 percent of bonds that do trade on a given day trade as much as four times that day, and there is only a one in three chance of repeating that on the following day. They also estimate that less than 30 percent of municipal bonds will trade in any given year.
All of the above factors contribute to the high cost of trading municipal bonds. The lack of liquidity and the lack of transparency results in the deck being stacked against the individual investor—from the perspective of a broker-dealer, the individual investor is ripe for exploitation. This leads to the following conclusion: Only investors who know with a virtual certainty that they will be able to hold individual municipal bonds they purchase until maturity should even consider buying them. And individuals acting on their own should only buy them in the primary market (at the initial offering).
To understand how bonds are traded we begin with an explanation of the bid-offer spread. The bid is the price at which you can sell a security, and the offer is the price you must pay to buy a security. The spread is the difference between the two prices. The size of the spread (the difference between the bid and the offer) is directly related to the riskiness of the bond from the dealer’s perspective— the more (less) liquid the trading is in a particular bond, the smaller (greater) will be the spread. The size of the spread is also related to the riskiness of the credit—the weaker (stronger) the credit rating, the greater (smaller) the spread. A third factor impacting the size of the spread is the price volatility of the security—the greater (lesser) the volatility, the larger (smaller) the spread.
There are two other terms that we must define that relate to pricing. A markup is the amount the broker-dealer adds to the wholesale offer price when selling to the investor. A markdown (or haircut) is the amount the dealer subtracts from the wholesale bid price when buying from an investor. The pricing on a particular bond in the wholesale (interdealer) market might be quoted, for example, as 100–101. That means that the dealer’s bid (the price at which he is willing to buy) is 100, and his offer (the price he is willing to sell) is 101. However, if the retail investor was interested in buying that same bond the broker-dealer might offer to sell the bond at a price of 103. The difference between the inter-dealer offer price of 101 and the retail offer price of 103 (2 points, or roughly 2 percent) reflects the dealer’s markup. Alter-natively, if the retail investor wished to sell that same bond the broker-dealer’s bid might be just 98. The difference between the interdealer bid price of 100 and the retail bid price of 98 is the dealer’s markdown. What determines the size of the markup and markdown? Several factors come into play.
The first factor impacting the size of the dealer markup or mark-down is the riskiness of the bond. Just as the riskiness of a bond impacts the size of the spread, it also impacts the size of the markup or markdown. The riskiness of a bond from a dealer’s perspective is related not only to its liquidity, but also to its price volatility. When a broker-dealer holds a bond in inventory he is taking price risk. Since long-term bonds are more risky than short-term bonds, the markup or markdown on longer-term bonds is likely to be greater— as compensation to the dealer for taking that incremental risk.
A second factor is the relative transparency of the market. The more opaque the market, the more the brokerage firm will be tempted to increase its profit on the trade. The reason is that the investor is unlikely to find out just how large a profit was taken— how much they were exploited. Because the market for Treasuries is so liquid, brokers have to be very careful when determining the size of their markup. However, when it comes to municipal bonds and mortgage-backed securities, the almost total lack of available pricing information provides a great temptation to widen spreads.
Another factor that determines the size of the spread is the size of the trade. For example, in the corporate market a $100,000 trade is considered to be a normal block size, while in the municipal market $50,000 would be considered to be a normal block. Trades that are smaller will have a larger markup or haircut (markdown) applied—the smaller the trade, the larger the impact on the price. Break points in terms of pricing can be expected at $1 million, $500,000, $250,000, $100,000, $50,000, $25,000, and below $25,000. An example will illustrate the impact of size on the cost of trading municipal bonds.
A municipal bond might exchange hands between dealers at 100 (i.e., at par). The dealer might offer to sell to a retail investor a block of $50,000 at a price of 102. If the buyer, however, was purchasing just $25,000 the offer price might now be 102.5. For an even smaller lot of just $10,000 the offer might be 103 or even 103.5. Conversely if the price were for a block of $250,000 the offer might be just 101.75, and for $500,000 the price might be 101.5. If another retail client wanted to sell that same bond to the dealer the bid for a block of $50,000 might be 98, dropping to as low as perhaps 97 or 96.5 for a lot of just $10,000.
Unfortunately, the MSRB provides almost no guidelines as to a limit on the amount a broker-dealer can mark the price of a bond up or down. The only guidance is what is known as Rule G-18. It requires that brokers trade at prices that are “fair and reasonable in relation to prevailing market conditions.” Given that the makeup of the MSRB member board is five representatives from bank dealers, five representatives from securities firms, and just five representatives from the public, it is easy to see that any over-sight is biased in favor of the industry, not the investor.
It is also unfortunate that the SEC does not require a broker-dealer to disclose the amount of markup or markdown charged. The result is that transactions costs in bond trades can be like icebergs, where the largest part (seven-eighths) is hidden beneath the surface. Because glacier ice is only slightly lighter than an equal amount of seawater, most of an iceberg remains below the waves. It is very dangerous for boats to travel in icy waters without equipment that can help them locate an iceberg.
When buying bonds, the danger is that because the only required disclosure is the transaction fee (which is not a commission, rather an administrative fee), most investors assume it is the only cost they will incur—typically a nominal amount such as $25 or $50. As we have seen, however, this is definitely not the case—it is just the tip of the iceberg. Without the assistance of the technology needed to identify markups and markdowns, investors may incur large costs of which they are not even aware.
Markups and markdowns can be very large, since there is little in the way of regulations to prevent abuses. We are sure the following will shock most investors.
In a May 2002 ruling, SEC administrative law Judge Lillian A. McEwen dismissed fraud charges brought by the SEC and the MSRB against a Los Angeles broker. McEwen concluded: “Markups and markdowns on municipal securities ranging from 1.87 to 5.64 percent were not excessive and did not violate the securities fraud laws.” 1 To put this in context, Vanguard’s bond funds carry operating expense ratios of only about 0.2 percent.
If those markups and markdowns don’t shock you, how about these? The Web site MunicipalBonds.com reports each quarter on the one hundred worst municipal bond trades. The worst trades are defined by the size of the spread between the price a customer sold a bond to a dealer and the price another customer paid the dealer for that same bond. The worst trade reported during the second quarter of 2004 was a spread of 80 percent. The spread on the one-hundredth worst trade was still in excess of 7 percent.
Although brokerage firms are legally allowed to charge undisclosed markups ranging upward of 5 percent, the practice is unfair because it takes advantage of investors who might not be aware that bonds commonly include markups from a broker-dealer (considering the markup charged may often be disproportionate to the riskiness of the bond’s features). Whether legal or not, fees of this size are certainly not in the interest of the investor, and the failure to disclose them is an indication that broker-dealers would agree.
A dealer is entitled to a markup or markdown for the services they provide. However, its size should reflect the risk entailed. For example, if a broker were asked to bid on a risky and illiquid bond, then he would be entitled to a greater markdown than would otherwise be the case. On the other hand, if he were simply going to immediately resell that same bond into the secondary market, then only a very small markup would be appropriate. Unfortunately, the size of the spread may provide more insight into the integrity of the dealer than it does on the risk of the transaction.
Fortunately for investors, some light is beginning to be shed on bond pricing. All municipal and corporate bonds, as well as all pass-through securities (such as mortgage-backed securities like GNMAs), are assigned a nine-digit CUSIP (Committee on Uniform Security Identification Process)—a means of identification. It is the equivalent of a stock symbol (all stocks also receive CUSIPs). CUSIPs allow market participants to track the price of bonds as they are bought and sold in both the wholesale (inter-dealer) and retail (investor) markets. Because all trades, and the prices at which they occurred, must be reported, those that have access to the technology can determine approximately what markup or markdown was added to the interdealer price. Bloomberg, a large financial media firm that provides financial data to participants in the securities markets, provides that service (as part of a package of related services) for a very substantial fee (about $20,000 per year). Investors can ask their broker-dealer to show them the prices at which the bond they bought or sold traded at on the day of their transaction (or the days surrounding the trade). Good luck getting that information. A more likely alternative is to find an unbiased (fee-only) investment advisor who can provide this information for you.
In the classic fairy tale “Snow White and the Seven Dwarfs,” the evil Queen, Snow White’s jealous stepmother, arrives at Snow White’s cottage disguised as an old peddler woman. Despite being warned by the seven dwarfs to not open the door for anyone or accept any gifts, Snow White answers the door. The Queen uses the girl’s naivete against her and lures Snow White into taking a bite from a poisoned apple. Falling into a sleeping death, Snow White can only be awakened by love’s first kiss.
The moral of the story is that children should be wary of old ladies who come knocking at their door tempting them with treats. It is more than likely that the old lady has a hidden agenda. The broker-dealer community knows that individual investors lack sufficient knowledge about the bond market, which makes exploiting them as easy as “taking candy from a baby.” Unfortunately for those investors, Prince Charming will not be riding in on a white horse to save them or their portfolio. The following are just two examples of how brokers exploit the naivete of investors.
The first example relates to the maturity of the bonds brokers prefer to sell. Investors need to be aware that the size of the impact of a markup or markdown on the yield of a bond is negatively related to its remaining term-to-maturity. For example, a bond with a remaining maturity of just one year will see a return impact of one percent for each one point of markup or mark-down. However, the impact on yield is reduced as the term-to-maturity lengthens because the markup will be “amortized” over a longer time. Consider the following example.
A bond with a remaining term of just one year is yielding 3 percent and trading at par (100). A markup of even 1 percent would be very hard to hide, as the yield-to-maturity would drop by more than 1 percent to 1.98 percent. On the other hand, if the bond had a remaining term of ten years, the yield-to-maturity would fall to about 2.85 percent. That is a drop in yield of just 0.15 percent. And if the bond had a remaining term of twenty-five years the yield-to-maturity would fall to about 2.93 percent. That is a drop in yield of just 0.07 percent. The longer the maturity, the less the impact on yield-to-maturity and the easier it is to hide the markup. Now imagine a broker wanting to take a markup of four points on the same bond. That would be very hard to do with a bond with just one year remaining to maturity because the yieldto-maturity would then be negative—the investor would pay 104 for a bond that in one year would return his $100 in principal and just $3 of interest. On the other hand, the impact on the yield-to-maturity of a bond with twenty-five years to maturity would be only about thirty-one basis points (0.31 percent) per annum.
It isn’t hard to guess which maturities brokers push when selling bonds to individual investors. Unfortunately, not only do investors end up paying large transaction fees, but they also end up taking far more price risk than would be typically prudent. (We will discuss the historical evidence on the relationship between risk and term-to-maturity in the next chapter.)
Another example of how brokers can exploit individual investors is by selling them premium bonds that have call dates that are much closer than the maturity date. The bonds sell at a premium as a result of their high coupons (the attraction) relative to current market rates. Again an example will illustrate how brokers can take advantage of the opaqueness of pricing and exploit an investor’s lack of knowledge about bond pricing and the risks of fixed-income investing.
A bond with a remaining term-to-maturity of twenty-five years is carrying a coupon of 6 percent. The current market rate for similar bonds is now 4 percent and the bond can be called at 101 in one year. Despite the relatively high coupon (2 percent above the current market rate) and the long remaining term-to-maturity, the bond should not be trading much above the call price of 101 because of the nearness of the call date and the high likelihood that the bond will be called by the issuer. Let’s assume the bond is trading at 103. Now the broker decides to add a markup of five points and sells the bond to the investor at 108. The broker might tell the investor that the yield-to-maturity is about 5.6 percent, well above current market rates. (Remember from our prior example how a long maturity can camouflage the size of the markup.) Unfortunately, the bond will almost certainly be called in one year. Assuming it is called at 101 the investor will have earned the coupon of 6 percent and lost seven points in price, producing a net loss of 1 percent. As you can see, the investor was actually sold a bond with a negative expected return! This is not as rare as you might think. But it gets worse. When the bond is called the broker will call the investor to advise him of the call. The investor now has to reinvest the cash and the broker gets to play the same game all over again.
A less extreme example would be if the call date were in three years instead of just one. In this case the price of the bond in the wholesale market would be around 106. With the same markup of five points the bond would be sold to the investor at 111. If the bond is called at 100 in three years, as is likely, then the investor would have earned a return of less than 3 percent (having earned the coupon of 6 percent but also having to amortize the premium of eleven points over just three years).
It is important to note that while the law does not require dealers to quote the yield-to-maturity, they are required to quote to the yield-to-worst. As you have seen, however, despite this requirement abuses occur. If you ever purchase an individual bond be sure that you ask the dealer to disclose both the yield-to-worst and the yield-to-maturity.
The expression “Beware of Greeks bearing gifts” is derived from the classical epic The Iliad, in which a large, hollow horse made of wood was used by the Greeks to defeat the Trojans. The resourceful Greeks hid soldiers inside the horse and left it outside the gates of Troy. They anchored their ships just out of sight of Troy and left a man behind to say that the goddess Athena would be pleased if the Trojans brought the horse inside the city and honored it. The Trojans took the bait, against the advice of Cassandra and Lao-coon. That night the Greek army returned to Troy. The men inside the horse emerged and opened the city gates for their companions. The Greeks sacked the city, thus winning the war.
The investment equivalent of the Trojan horse is a broker bearing a gift in the form of a great bond for you to buy. A broker calls a customer, or potential customer, and says: “We have a unique opportunity to buy this great bond. It is a real bargain. I have to tell you that it won’t be available for long.” The broker concludes, “How much can I put you down for?”
If you ever hear a pitch like this, the first and only response you should make is to hang up. First, there are no great bargains in the bond market. Second, even if such a bargain existed you should ask yourself, “Why are they offering to sell it to me instead of to some large institutional client with whom they do millions, if not hundreds of millions, of dollars in business every year?” In all likelihood, the reality is that the trading desk of the brokerage firm wants to unload some bond that they no longer want to hold in inventory (and possibly for a very good reason, like the credit is deteriorating or the market is looking weak for similar securities). In order to move the bond as quickly as possible the trading desk will offer the brokers of the firm a larger concession (price discount to which the broker can then apply a markup) as an incentive to sell the bond. Typically the broker would then add his or her own markup to further increase his or her compensation. Great deal for the broker, lousy deal for the unsuspecting investor. Remember that when a broker says we have a great opportunity to buy a bond what he is really saying is that he has a great opportunity to sell a bond. You are doing the buying, not the broker!
One last point: In her book The Bond Bible, veteran money manager Marilyn Cohen recommends that whenever a broker calls with a recommendation on a bond and tells you that he just bought that security for himself (or his parents) be sure to ask for a copy of that confirmation slip. She describes this tactic as one of the most popular fibs. The same advice applies to recommendations to buy stocks or mutual funds. Whenever you get advice from brokers or financial advisors you should always ask to see their financial statement to see if they are investing in the same securities they are advising you to purchase—making sure that they are “putting their money where their mouth is.”
The above examples illustrate how essential it is to be an educated investor. Hopefully they also have shown you that while education doesn’t have to be expensive (consider the price of this book), ignorance can be very expensive when you have brokers who don’t have their customers’ best interests at heart.
We have now completed our discussion of how bonds are bought and sold. The next chapter focuses on how the academic community believes the fixed-income markets really work—how markets set security prices. We will also explore the relationship between risk and reward. Keep in mind that most investors obtain their “knowledge” of how markets work from Wall Street and the financial media. This is, unfortunately, the equivalent of getting medical advice from People magazine. Learning how the academic community believes markets work, on the other hand, is the equivalent of getting medical advice from the rigorously peer-reviewed New England Journal of Medicine. The knowledge we gain will lead us to the winning investment strategy.