iStock_000016105828_XSmallA stale pricing edge occurs when a security or fund can be purchased or sold at a price that is stale with respect to current up-to-the-second information. This trading edge is as fleeting as a twenty dollar bill sitting on a busy sidewalk. For instance, if the price of a U.S. equity closed-end fund is sitting at bid $10, ask $10.10 for most of the day without trading activity, and the S&P 500 trades up 2% during the day, the ask price of $10.10 may be too low, and therefore, stale. Then buying at the bid can provide a risk-free profit (if hedged by an S&P 500 short) of at least 10 cents, since the fund should be trading at $10.20/$10.30.

You’ll never find stale prices with liquid large cap equities or ETFs. Only rarely do they occur with securities that don’t trade very often, such as closed-end funds. The stale price edge is not backtestable with price data since simulated orders would have affected prices at the time.

Generally, we’d expect stale price opportunities to occasionally occur with illiquid stocks that trade with a wide bid-ask spread, with very little size offered. Stale prices may also emerge when transaction costs are high (perhaps due to a financial transaction tax), or when markets are highly volatile. In the heat of a stock market crash, when 10% up and down days are the norm, we’d expect a few stale price opportunities to emerge. Getting back to the U.S. equity closed-end fund discussed above, if the S&P 500 ramped 5% in the last hour of the day, and the closed-end fund lagged up only 2%, then perhaps that’s an opportunity to buy at the ask, and expect its price to eventually catch up (possibly the next morning).

Funds are the bread-and-butter tools for asset class traders – ETFs, ETNs, closed-end funds, open-end funds, limited partnerships, etc. There are stale pricing opportunities that can occur with investment funds from time to time.

Most notable was the stale pricing associated with open-end mutual funds. This advantage is no longer available today, but it represented a structural edge from the 1980s to the early 2000s. In his book The New Market Wizards, Jack Schwager profiled one trader, Gil Blake,1 who used this edge in the 1980s to produce an annualized return of 45% per year over 12 years. Mr. Blake actively traded Fidelity sector funds to generate those returns, switching in and out of funds every day. Interestingly, he was unaware of why the strategy worked, when what he was doing was systematically exploiting stale pricing. At that time, mutual fund prices exhibited a daily serial correlation because many of the securities held in the fund had stale prices when used to calculate the daily close-of-business net asset value (NAV).

New York Attorney General Eliot Spitzer put a stop to this sort of trading in the early 2000s,in what is known as the mutual fund market timing scandal.2 Spitzer sued many large mutual fund companies for allowing favored clients to perform this sort of active in and out trading at the expense of long-term shareholders. After the scandal, mutual fund companies moved to monitor and restrict this sort of trading.

Spitzer also charged a few hedge funds engaging in an illegal form of stale-price trading called “late trading.” Late trading allowed a trader to submit a mutual fund trade after the close (in reaction to an after-hours earnings release) that was effective as of the close at 4:00 pm EST.2 Prior to 1968, mutual fund investors could actually purchase and sell mutual funds at the previous day’s closing price. This allowed a guaranteed profit until the SEC cracked-down on this practice.3

But it was the introduction of decimalization in the early 2000s that dramatically reduced the stale price edge. To illustrate the power of this edge, I’ll examine a very simple and effective stale-pricing trading strategy. If the stock market is up for the day as of 5 to 10 minutes before the close, trade into a small cap equity fund before the close. If the stock market is down for the day, trade out of the small cap fund into a money market account. I’ve simulated this trading system using the T. Rowe Price New Horizons Fund (Symbol: PRNHX) as a proxy for the market (the Russell 2000 index could also be used), and three-month T-bill returns as a proxy for the money market account.

Note that such trading was not allowed using this fund during this time, with T. Rowe Price limiting trading at four round trips per year. I’m using this fund to illustrate the stale price effect due to its long history of price data. But as Spitzer revealed in September 2003, many hedge funds and other trading entities had secret arrangements with mutual fund families that allowed this sort of trading.

Table 1 shows the year-by-year results of such a strategy from 1984 to 2015. Also shown are the arithmetic averages over two periods. The first period, from 1984 to 2001, shows a strategy that produced an average return of almost 48% per year. Decimalization of stock prices began in mid-2001. Before decimalization, stocks traded in eighths or sixteenths, which for low priced and/or illiquid small cap stocks led to many days with unchanged and thus stale prices. In the mid-1980s over 50% of stocks had unchanged prices on a given day. That fraction drifted down over time until it was less than 10% around 2001. Once decimalization was incorporated, stock prices could incorporate market information much more efficiently and fairly.

In the second time frame, post-decimalization, we see that the benefit of this edge was completely eliminated, with the nominal non-risk-adjusted alpha shifting from an average of 34.5% to -3.1%. Also, the number of one-way trades increased from 101.1 per year to 124 per year, which is basically a 50:50 ratio with about 252 trading days per year.

Table 1: Simulated stale price trading with the T. Rowe Price New Horizons Fund

Year Buy & Hold Timing Model Nominal  Alpha One-Way  Trades


-9.5% 34.1% 43.6% 103
1985 24.2% 47.6% 23.4% 92
1986 -0.1% 30.1% 30.2% 93


-7.4% 63.6% 70.9% 101
1988 14.0% 41.7% 27.7%


1989 26.2% 37.7% 11.5%


1990 -9.6% 65.4%




52.3% 85.4% 33.1% 101
1992 10.6%




1993 22.0% 55.3% 33.2%


1994 0.3% 29.2% 28.9%



55.4% 69.1% 13.7% 97
1996 17.0% 45.6% 28.5%



9.8% 54.3% 44.5% 100
1998 6.2% 70.8% 64.6%



32.5% 27.4% -5.1% 117
2000 -1.9% 30.4% 32.3%



-2.8% 35.9% 38.8% 112
2002 -26.6% -19.4% 7.2%


2003 49.3% 27.2% -22.1% 129
2004 17.9% 22.3% 4.4%



11.9% 5.5% -6.4% 133


7.4% 16.5% 9.1%


2007 6.9% -0.1% -7.0%


2008 -40.9% -21.4% 19.5% 128
2009 49.6% 21.3% -28.3%


2010 34.3% 28.2% -6.1% 117
2011 6.1% 12.0% 5.9%


2012 13.8% 15.0% 1.1% 124
2013 51.2% 33.5% -17.7%


2014 6.9% 4.2% -2.7% 127
2015 5.1% 3.6% -1.5%


Average (1984 – 2001):


47.8% 34.5%


Average (2002 – 2015):

13.8% 10.6% -3.2%



Stale pricing with international funds, sometimes called time-zone arbitrage,4 was also an edge exploited by fund traders. In this case, fund traders can buy and sell international equity funds at a stale price (based on when European and Asian stock markets closed and new information associated with how the S&P 500 closed at 4:00 pm EST). The simple trading system is to buy an international equity mutual fund when the S&P 500 is up for the day (as of 5 to 10 minutes before the close), and shift to a money market account when the S&P 500 is down for the day. Table 2 shows the simulated results of such a strategy.

Table 2: Simulated stale price trading with the T. Rowe Price International Stock Fund

Year Buy & Hold Timing Model Nominal  Alpha One-Way  Trades


31.6% 40.2% 8.6% 120
1990 -3.2% 39.9% 43.1%



30.4% 44.7% 14.3% 124
1992 7.6% 21.1% 13.5%



9.7% 33.1% 23.4% 125
1994 0.4% 36.5% 36.1%



38.0% 36.2% -1.8% 129
1996 22.6% 37.2% 14.6%



33.5% 62.8% 29.4% 131
1998 28.7% 48.8% 20.1%



20.8% 68.7% 47.9% 118
2000 -9.7% 63.0% 72.7%



-11.8% 40.2% 52.0% 131
2002 -21.6% 48.8% 70.4%



28.2% 79.8% 51.6% 132
2004 10.7% 20.8% 10.1%



4.8% 15.0% 10.1% 134
2006 15.8% 29.5% 13.6%



5.1% 2.8% -2.3% 144
2008 -36.8% -36.7% 0.1%



26.4% 15.7% -10.6% 125
2010 14.6% -0.6% -15.2%



1.9% 4.7% 2.8% 119
2012 16.0% 14.2% -1.8%



32.3% 15.8% -16.5% 130
2014 13.5% -1.3% -14.8%



-0.8% 9.7% 10.4%


Average (1989 – 2003):


46.7% 33.1%


Average (2004 – 2015):


7.5% -1.2% 126.9

Spitzer shut down this sort of trading in the fall of 2003, which is the date I use to demark when the edge was eliminated. This stale pricing edge would have produced an arithmetic return of 47% per year from 1989 to 2003. After that, mutual fund companies implemented a number of ways to thwart this sort of trading, including the implementation of a short-term redemption fee (such as a 2% fee for any trades not held for 30 days), and instituting a fair-value pricing approach to account for stale prices. After these changes, the effectiveness of the strategy was eliminated. International ETFs do not have this effect because they trade throughout the day and account for all new information from when foreign markets close and the U.S. market closes.

Future Opportunities

Are there opportunities like this anymore? No, not with these sorts of theoretical returns. But there are funds that can provide stale-price trading opportunities that are economically meaningful. Fund auditors have a system for characterizing the accuracy of security pricing, ranking securities from Level 1 (highly liquid) to Level 3 (highly illiquid). Funds that hold Level 1 securities that actively trade throughout the day will not have stale prices. Funds that hold Level 1 securities that are priced each day, but don’t trade very often due to wide bid/ask spreads and high financial transaction taxes and costs, may have a possibility of stale pricing. Markets with a large fraction of securities with unchanged prices each day may have exploitable stale prices.

Funds that hold illiquid Level 2 and Level 3 securities will quite possibly have some stale pricing associated with them. Such funds may hold private equity, illiquid bonds, real estate, private contracts, etc. Funds that use model-based or appraisal-based pricing5 or have other pricing quirks6 will likely have stale pricing. Money market account arbitrage is one example.7

There are always fund managers tempted to invest too much assets in illiquid holdings packaged in a 1940-act mutual fund. The Third Avenue Focused Credit Fund (TFCIX) is the latest example.8 Hedge fund managers who manage limited partnerships that hold illiquid holdings, but offer monthly liquidity to investors, can provide a stale price edge around entry and exit decisions. While these limited partnerships do not allow for daily in and out trading, they do provide a costless option at the entry and exit point of investment.

Interval funds holding illiquid securities, such as the SharesPost 100 Fund (Symbol: PRIVX), is another example. This fund holds private equity in late-stage, venture-backed companies, primarily based in the Silicon Valley. Such funds can provide a stale price free option-like benefit at the entry and exit point of an investment.

When investing in limited partnerships, such as with real estate, sometimes you’re given the opportunity to invest with more information than initial investors if the funding window hasn’t closed.

Of course, selling out of a fund when you suspect prices are stale, or when you expect a run on the fund, can provide a stale price edge.9,10,11

Wall Street is constantly developing new financial products to sell to unsuspecting investors. At times, these developers are young and overzealous and don’t realize they’ve created product that is poorly constructed. I always investigate any new product for potential stale pricing.

Stale Pricing in Trend-Following Backtesting

All simulated index data has stale pricing. When actual tradable securities based on the index are introduced, the stale pricing effect is eliminated. For this reason, great care is needed when interpreting trend-following, backtested results on non-tradable indexes. Some backtesters like to evaluate systems on Dow Jones data going back to the 1920s or S&P 500 daily data going back to the 1950s. Backtested trend-following performance is probably overstated due to stale pricing. I don’t trust any of this backtesting. As I discuss in Trend-Following Market Timing, I only trust data going back to the mid-1980s because that’s when S&P 500 futures were introduced. Be very careful about employing a new trend-following system based on simulated data motivated by the introduction of a new trading vehicle.

Stale Pricing in Premiums and Discounts

When evaluating trading opportunities derived from ETF, ETN, and closed-end fund (CEF) premiums and discounts, you must account for stale price effects in the quoted NAVs. Of course, international ETFs have quoted NAVs that are stale every day. Many CEFs update NAVs on a weekly basis and/or have some private equity holdings that do not price every day. CEFs that trade at a perpetual premium or large discount must be investigated for stale pricing before implementing a mean-reversion trade associated with the narrowing of a premium or discount.

At times, an illiquid CEF can have a head-scratching daily move that seems completely counter to what’s happening in the broader market, but in reality is just correcting a stale price from the previous day. Foreign markets can close for extended periods of time, which can cause large discrepancies between quoted ETF NAVs and prices. Past data for extremely illiquid ETFs can imply backtested trading opportunities, but what some don’t realize is that historical prices were not available at the time because the bid/ask prices properly accounted for the actual NAV at the time.


Be on the lookout for stale pricing moments. They don’t occur very often, but when you find an opportunity, it constitutes a very solid and exploitable trading edge. If the edge is structural in how the trading vehicle is constructed, then a systematic trading approach can be developed to exploit the mispricing.



  1. Schwager, J.D., The New Market Wizards: Conversations with America’s Top Traders, 1992, pp. 230-250.
  2. 2003 mutual fund scandal, Wikipedia.
  3. Pricing of Redeemable Securities for Distribution, Redemption and Repurchase and Time-Stamping of Orders by Dealers, Securities and Exchange Commission, 1968.
  4. Warwick, B., Searching for ALPHA: The Quest for Exceptional Investment Performance, 2000.
  5. Redding, L.S., “Persistent Mispricing in Mutual Funds: The Case of Real Estate”, Working Paper, March 2005.
  6. McCrum, D., “Meet the man who could own Aviva France”, Financial Times, February 27, 2015.
  7. Crescenzi, T., “Money Supply Takes a Big Plunge”,, July 19, 2004.
  8. Zuckerman, G., and Wirz, M., “How the Third Avenue Fund Melted Down”, Wall Street Journal, December 23, 2015.
  9. Pulliam, S., Smith, R., and Siconolfi, M., “U.S. Investors Face an Age of Murky Pricing”, Wall Street Journal, October 12, 2007.
  10. Goldstein, M., Henry, D., and Der Hovanesian, M., “The Subprime Blowback: Mutually Assured Mayhem”, Bloomberg Businessweek, July 9 and 16, 2007.
  11. Asia Insight, Matthews Asia Funds, April 2005. Short description of Tawainese bond mutual funds that suffered heavy redemptions at prices above NAV.


The content contained within this blog reflects the personal views and opinions of Dennis Tilley, and not necessarily those of Merriman Wealth Management, LLC. This website is for educational and/or entertainment purposes only. Use this information at your own risk, and the content should not be considered legal, tax or investment advice. The views contained in this blog may change at any time without notice, and may be inappropriate for an individual’s investment portfolio. There is no guarantee that securities and/or the techniques mentioned in this blog will make money or enhance risk-adjusted returns. The information contained in this blog may use views, estimates, assumptions, facts and information from other sources that are believed to be accurate and reliable as of the date of each blog entry. The hypothetical data displayed in this blog does not represent actual performance and should not be interpreted as an indication of actual performance. Although we have done our best to present this information fairly, hypothetical performance is still potentially misleading. This data is based on transactions that were not made. Instead, the trades were simulated, based on knowledge that was available only after the fact and thus with the benefit of hindsight. Results do not include the impact of taxes, if any.  The content provided within this blog is the property of Dennis Tilley & Merriman Wealth Management, LLC (“Merriman”). For more details, see the Important Disclosure.”