What is the most effective government policy to boost stock returns and, subsequently, to foster economic growth, especially during an economic crisis? This paper examines the effects of various government actions on stock market performance during the global financial crisis of 2007. As expected from rational investors, nearly all of the variation in index levels is explained by the variation in index futures levels, which reflect market expectations. Index and future levels were collected for June 2007-December 2009 using Bloomberg, which tracks data for various financial instruments. The financial crisis timeline furnished by the New York Federal Reserve Board on its website was used to determine important policies, which were then categorized into eight groups. Dummy variables were created for each of these categories corresponding to the date on which the event in question occurred. VIX futures are used to model market expectations of volatility; federal funds futures, of the federal funds rate. Regressions using the ordinary least squares measure and correcting for autocorrelation were performed with market expectations of index levels as the dependent variable. Significant predictors of index futures were volatility futures, federal funds rate futures, and the creation or adjustment of swaps. The most effective way to change stock market returns is to change expectations, and the most practical way to change expectations is to decrease the federal funds rate.
Financial crises are a fact of life in modern economies. Government actions to restore the economy are also all but inevitable. While there are multiple ways to quantify the health of modern economies, among the most visible are stock markets and their representative indices. Examining the reactions of stock market indices is then a quite natural experiment to measure the effects of various government actions undertaken to revitalize the economy. The global financial crisis of 2007-2009 provides us with a rich natural experiment to do just that. While many governments intervened in an effort to temper the effects of the crisis, we study just the United States market because it has readily available data, something scarce in examining other countries. Although the markets have rallied in response to most government growth initiatives, assaying which initiatives are most successful may allow us to better handle future crises.
To construct an appropriate model, we must examine its components and the theoretical foundations that undergird their inclusion in the model. The following sections detail the results of this examination.
Why are stock market returns a valid measure of the effects of government actions on investor actions and subsequent economic effects? Malcolm Baker and Jeffrey Wurgler found that for hard-to-value stocks, pessimism causes lower valuations and optimism causes higher valuations than would exist in the absence of sentiment. Ergo, although stock market returns will have greater variance than investor sentiment, they still reflect changes in investor sentiments and in fact magnify these changes. Baker and Wurgler also identified high volatility stocks as a category that is hard to value and therefore subject to the effects of investor sentiment. Large swings in major stock indices show that high volatility permeated the United States stock market during the 2007 crisis. Over half of the Dow Jones Industrial Average’s twenty largest daily losses and twenty largest daily gains ever occurred in 2008.The Standard and Poor 500 shows a similar trend: Approximately half of the twenty largest daily gains and twenty largest daily losses occurred in 2008. In regards to the economic effects, Pearce has shown that returns Granger cause economic growth in the United States. Higher returns, then, are a manifestation of policy successes, even if investor sentiment magnifies them.
Expectations are included in our model on the basis of two major theories in economics: the Efficient Markets Hypothesis and the Lucas Critique. The Efficient Markets Hypothesis states that “[a] market is efficient with regard to information set θt if zero economic profits [risk adjusted returns net of costs] can be made by trading on the basis of information set θt”. In other words, prices in an efficient market already reflect all information in a particular information set. While there are several definitions of the information set for which markets are efficient, the most widely used form of this hypothesis defines it as all publicly available information. The Lucas Critique asserts that the effects of policy decision are affected by people’s expectations of those policy decisions. Thus, if people believe that the government will enact some policy, the effects of the policy will not be as strong as people have already made decisions based on this belief.
Futures are used as proxies for numerical data. Futures are standardized, exchange-traded contracts that oblige the purchase or sale of an underlying asset on a specified future date. They are traditionally used for hedging (guarding against future uncertainty in cash flows) and speculating (investing based on beliefs about the future). Profits depend on whether the value of the underlying asset increases or decreases in the future. Ergo, futures markets reflect investors’ beliefs regarding the future of the market. Although investors do form expectations about policy events, it is more difficult to determine these expectations. However, if investor’s expectations regarding policy actions are right, policy effects will be tempered according to the Lucas Critique. Thus, our model’s estimates of policy effects, which take expectations into account, are lower than they would be had we found a way to model policy expectations.
Interest Rate Changes
Decreases in interest rates tend to encourage economic growth as the effective costs for both capital investments and consumer goods falls. The IS-LM model, based on the work of John Maynard Keynes, explains this relationship. National output is the sum of consumption, investment, and government spending. The equilibrium interest rate is determined by intersection of two curves graphed against national output: the downward-sloping investment and savings curve (the “IS” curve) and the upward-sloping liquidity and money curve (the “LM” curve). The former represents the market for goods and services and is affected by fiscal policies such as changes in government expenditures and tax rates. It slopes down because decreasing interest rates increase investment, which increases national output. The latter represents the money market and is affected by monetary policy, especially the supply of money in the economy, as determined by Federal Reserve policies. It slopes up because increased consumption, which leads to increased national output, increases demand for cash and increases the interest rate people will pay for cash. The intersection of these two curves represents the equilibrium interest rate for fixed level of government spending, taxes, money supply, and price level. Changing any of these will shift at least one of the curves and therefore change the prevailing interest rate. It is changed most often by altering the money supply. Below is a graphical representation of this model.
FIGURE 1: Graphical Representation of IS-LM Model
The Federal Reserve implemented several different policies to control interest rates, mostly lowering them (easing them). Mishkin argues that monetary policy easing is a potent tool to use during a crisis because it mitigates the effects of adverse feedback loops. Adverse feedback loops in this context are vicious cycles in which assets lose value, which makes them less useful as collateral or investments, making their value fall even further. Thus, easing monetary policy, which prevents large losses in asset value, should strongly impact the economy.
Open market operations are one of the Fed’s most beloved tools, even during expansionary times. Open market operations are traditionally used to set the federal funds rate, the rate at which depository institutions lend extra reserves to each other. By buying and selling federal securities from these depository institutions, the Federal Reserve changes their reserves and thereby affects the rate at which they will lend to other institutions, the federal funds rate, which determines many other interest rates. The Federal Reserve, which is also the “lender of last resort”, may also lower the discount rate, the rate at which the Fed itself lends to depository institutions. Although this rate is set above the target federal funds rate, lowering it eases liquidity crunches when banks cannot borrow funds from other banks. Even though banks no longer need to exhaust other sources of funding before borrowing from the Federal Reserve, it is cheaper as the federal funds rate is generally lower than discount window rates.
Market Stability Actions
For our purposes, a market stabilization action is any sort of capital infusion from either the Federal Reserve or the Department of Treasury. The goal of these programs was to stabilize financial markets. Under this category, we include three policy actions: bailouts, liquidity programs, and guarantees. The Federal Reserve extended loans to several companies, including J.P. Morgan Chase and American International Group (AIG).  Furthermore, the Department of Treasury bought “troubled assets” from hundreds of financial institutions in a program known as the Troubled Asset Relief Program. While experts were initially skeptical about this program, it has been successful: In addition to increasing credit availability, it has even been profitable.  The Treasury also invested in AIG, Chrysler, General Motors, and mortgage servicers and implemented a host of credit programs that allowed banks to borrow capital. The news media highly publicized these market stabilization actions, especially the large-scale bailouts. Because these actions were a major pillar of the United States Government plan to handle the financial crisis and involved large-scale purchases and sales of equities, bailouts should affect the stock market, as they displayed the government’s commitment to protect America from financial destruction.
To ease liquidity pressures in foreign markets, the Federal Reserve also opened “swap lines” or reciprocal currency arrangements with several central banks. In these reciprocal currency arrangements, the foreign bank sold a given amount of its currency in exchange for dollars with the promise to buy it back with interest at the same rate later. The foreign banks then lent the dollars out to depository institutions within their jurisdiction. Because these depository institutions had balance sheet items denominated in dollars but cash reserves in other currencies, they often had to borrow dollars to complete transactions. While borrowing dollars is generally a straightforward process, during the financial crisis, illiquid markets precluded this usual state of affairs during the financial crisis, leaving foreign depository institutions without access to the dollars they needed to carry out their transactions. At its peak, foreign currency from outstanding swap lines accounted for 25% of the Federal Reserve’s total assets, demonstrating the extent of the illiquidity in foreign countries.
Swap lines allowed these institutions to access dollars at reasonable interest rates. However, even the swap lines could not keep up with European (and other foreign) demand for dollars: High bid-to-cover ratios, which indicate high demand, dominated the auctions for these funds. Although the central banks do have foreign currency reserves, they were not enough to cover the demand for dollars. Moreover, the central banks would have crowded out and further crippled foreign depository institutions who did have dollars to lend by lending dollars at a better interest rate. Swap lines let American depository institutions and other foreign depository institutions keep their dollars instead of lending them out, strengthening banks worldwide. Opening swap lines successfully eased dollar lending pressures according to several measures.
VIX: The Fear Index
The CBOE Volatility Index, commonly known as the fear index, is the premier measure of expected volatility stock market volatility over the next 30 days. The Chicago Board of Exchange uses the premiums of many different S&P 500 options in its calculations of this index. The prices of these options reveal the market’s expectations of volatility as options more likely to be in-the-money (to have a positive payoff) command higher prices.  Options allow, but do not oblige, buyers to purchase or sell an underlying asset at specified date at a specified price. Thus, their payoffs and consequently their prices are based on the prices of the underlying assets and expectations thereof. Higher VIX prices are associated with higher expected market volatility. People will only engage in options trades that they believe will lead to profit; thus, the prices of options reflect peoples’ willingness to engage in such trades, meaning their belief in the profit potential. An option less likely to be in-the-money by its expiry date—less likely to be exercisable for a profit—will have lower price so that those who do buy the option can still make a profit off it: Otherwise, they would not sell. The likelihood of an option at a particular strike price (price at which the option is exercised) depends on the volatility of the underlying asset. Thus, options based on high volatility assets will have a higher likelihood of being in-the-money before expiry, meaning that people will be willing to pay more for those options as they can still expect to make a profit at the higher price. Therefore, the VIX Index, which tracks options prices, also tracks the volatility of the underlying asset, the S&P 500 Index. The VIX reflects fear of a downward shift in the market more than exuberance due to an upward shift because the futures market consists largely of loss-averse hedgers insuring their assets against steep losses. Low further articulates this point by showing that price decreases lead to VIX increases faster than price increases lead to VIX decreases.
As would be expected, events involving private companies (“market events”) play a large role in the performance of the market, especially when they involve well-known companies or large swaths of the market. The Federal Reserve Bank of New York (FRBNY) even includes major market events in its review of the financial crisis to further illuminate policies adopted by various federal agencies. Our model takes into account the events the FRBNY found significant enough to include in its review.
Government-sponsored enterprises (GSEs), such as Freddie Mac and Fannie Mae, are private companies “established and chartered for public policy purposes”. Both the government and private investors hold stock in these companies. They primarily serve people and businesses that would otherwise be neglected. For our analysis, we separate events related to GSEs because GSEs’ dual nature as private companies with public policy missions make it difficult to classify these events as private market events or as policy actions.
Policy actions are announcements by government officials or policies undertaken by groups other than the Federal Reserve or the Treasury. These include the White House, the Securities and Exchange Commission, and Congress. Although we expect that these policies will have an effect on the market, as they often involve market participants, they do not belong in any of the other categories.
The model consists of three main parts. The left hand side variable is Standard and Poor’s 500 stock index futures. As expected from rational investors, almost all of the variation in stock price, as measured by Standard and Poor’s 500 stock index, is explained by the variation in expectations, the futures. Thus, rather than using the stock prices themselves, futures are used to determine how events affect expectations. On the right hand side, the first part is government policy actions (as well as market events and GSE events). These have been divided into the following categories: swaps, bailouts, spreads, policy actions, market stabilization events, reserve interest, foreign bank interest, and market events. The justifications for these divisions have been explained in the preceding section on theoretical foundations. We model these variables using dummies for the days on which particular events happened. The second is numerical data: VIX index and futures data and federal funds futures data. For the futures, used to model expectations, we use the price of the contract closest to expiry on that date for expediency, as there is no contract that covers the entire period studied. VIX and federal funds are monthly contracts, while the S&P 500 is a quarterly contract. All numerical data was downloaded from Bloomberg while all policy data was collected from the Federal Reserve Bank of New York Financial Turmoil Timeline. We estimate parameters using both regular ordinary least squares procedures as well as procedures correcting for autocorrelation with lags of one day, one month, one quarter, and one year to correct for any possible daily effects, seasonal effects, and calendar effects.
Results and Analysis
Presented in the tables below are the results of our examination, both the parameter estimates and the corresponding standard errors. Results for both ordinary least squares (OLS) and autocorrelation-corrected regressions are shown. Although both Durbin-Watson test statistics and Bruesch-Godfrey LM test statistics were significant at the 5% level, the results for ordinary least squares and autocorrelation-corrected regressions were very similar. Single asterisks (*) represent significance at the 10% level while double asterisks (**) represent significance at the 5% level. Henceforth in our discussion, significance will mean significance at the 5% level unless otherwise specified.
FIGURE 2: Parameter Estimates
FIGURE 3: Standard Errors
Because futures prices, a proxy for expectations, so strongly influence stock prices, we consider which policy actions and numerical data affect investor expectations. VIX futures and federal funds futures are significant. The index futures price falls by eleven points for every dollar increase in VIX futures, reflecting imminent uncertainty and risk, while it plummets by 82 points for every dollar increase in the federal funds futures price, reflecting the negative effects of higher expected interest rates. The results of using expectations lagged by one day are very similar. The creation or adjustment of a currency swap is also significant at the 10% level with a parameter estimate of 39. When corrected for autocorrelation, it is also significant at the 5% level with a slightly higher parameter estimate of 43. One possible explanation for this is that currency swaps stabilize the foreign exchange markets, making the relatively volatile stock market seem more attractive in terms of returns.
Several avenues exist for future research. While policy actions were modeled as dummy variables, many of them have numerical values, such as a certain amount of money or a particular percentage change. Incorporating these values would shed further light on what is effective and to what extent. Examining lagged data (data from previous days) would also be interesting as it could reveal the effects of prior data on current data; however, the appropriate number of days to backtrack is still to be determined. Lastly, one could also examine data from other sovereigns, many of whom used similar measures.
During the global financial crisis of 2008, the United States government performed many actions to stabilize and strengthen the economy. While it is not by any means the only measure, stock market activity is one of the primary measures used in America when discussing the economy. Shifts in the stock market are predicated upon expectations as revealed in the futures market, especially the index futures, VIX futures, and federal funds rate futures. When we consider what influences investor expectations as measured by Standard and Poor’s 500 stock index futures, prices for other futures, such as the VIX and the federal funds futures rate are significant. In fact, the only policy action that significantly affects index futures prices is the creation or the adjustment of swaps. Thus, the most effective way to intervene in the economy such that the stock market reacts positively is to change expectations, a rather difficult endeavor. Decreasing the federal funds rate appears to be the most practical way for the government to change expectations and thereby increase stock prices.
“Bailout: Will It Work?” CNNMoney, October 4, 2008. http://money.cnn.com/2008/10/04/news/economy/will_it_work/index.htm.
Baker, Malcolm, and Jeffrey Wurgler. “Investor Sentiment in the Stock Market.” The Journal of Economic Perspectives 21, no. 2 (April 1, 2007): 129–151. doi:10.2307/30033721.
Barr, Colin. “TARP Profit Claim Bugs Skeptic.” CNNMoney, August 4, 2009. http://money.cnn.com/2009/08/04/news/economy/tarp.profit.fortune/.
Board of Governors of the Federal Reserve System. “Credit and Liquidity Programs and the Balance Sheet.” Federal Reserve, July 30, 2012. http://www.federalreserve.gov/monetarypolicy/bst_lendingdepository.htm.
“Budget of the United States Government, Fiscal Year 2013.” Office of Management and Budget, 2010. http://www.whitehouse.gov/sites/default/files/omb/budget/fy2013/assets/budget.pdf.
“Dow Jones Industrial Average All-Time Largest One Day Gains and Losses.” Wall Street Journal, December 31, 2012. http://online.wsj.com/mdc/public/page/2_3047-djia_alltime.html.
Federal Reserve Bank of New York. “Open Market Operations.” Federal Reserve Bank of New York, August 2007. http://www.newyorkfed.org/aboutthefed/fedpoint/fed32.html.
“Financial Turmoil Timeline.” Federal Reserve Bank of New York, December 31, 2010. http://www.newyorkfed.org/research/global_economy/Crisis_Timeline.pdf.
Fleming, Michael J., and Nicholas J. Klagge. “The Federal Reserve’s Foreign Exchange Swap Lines.” Current Issues in Economics and Finance, April 2010.
Jensen, Michael. Some Anomalous Evidence Regarding Market Efficiency. SSRN Scholarly Paper. Rochester, NY: Social Science Research Network, August 12, 2002. http://papers.ssrn.com/abstract=244159.
Low, Cheekiat. “The Fear and Exuberance from Implied Volatility of S&P 100 Index Options.” The Journal of Business 77, no. 3 (July 2004): 527–546.
Mankiw, N. Gregory. Macroeconomics. 7th ed. New York: Worth, 2009.
McDonald, Robert. Fundamentals of Derivatives Markets. Upper Saddle River, NJ: Prentice Hall, 2008.
Mishkin, Frederic. “Is Monetary Policy Effective During Financial Crises?” NBER, January 2009. http://www.nber.org/papers/w14678.pdf?new_window=1.
Pearce, Douglas. “Stock Prices and the Economy.” Economic Review (November 1983): 7–22.
“S&P 500 Index All-Time Largest One Day Gains and Losses.” Wall Street Journal, December 31, 2011. http://online.wsj.com/mdc/public/page/2_3047-sp_alltime.html.
Schwabish, Johnathan, and Courtney Griffith. Troubled Asset Relief Program. Web, March 28, 2012. http://www.cbo.gov/sites/default/files/cbofiles/images/pubs-images/43xxx/TARP_3_27_2012.png.
“The CBOE Volatality Index.” Chicago Board of Exchange, 2009.
Whaley, Robert. “Understanding VIX,” November 6, 2008. http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1296743.
 Malcolm Baker and Jeffrey Wurgler, “Investor Sentiment in the Stock Market,” The Journal of Economic Perspectives 21, no. 2 (April 1, 2007): 130, doi:10.2307/30033721.
 “Dow Jones Industrial Average All-Time Largest One Day Gains and Losses,” Wall Street Journal, December 31, 2012, http://online.wsj.com/mdc/public/page/2_3047-djia_alltime.html.
 “S&P 500 Index All-Time Largest One Day Gains and Losses,” Wall Street Journal, December 31, 2011, http://online.wsj.com/mdc/public/page/2_3047-sp_alltime.html.
 Douglas Pearce, “Stock Prices and the Economy,” Economic Review, November 1983, http://www.kc.frb.org/publicat/econrev/EconRevArchive/1983/4q83pear.pdf.
 Michael Jensen, Some Anomalous Evidence Regarding Market Efficiency, SSRN Scholarly Paper (Rochester, NY: Social Science Research Network, August 12, 2002), 3,7, http://papers.ssrn.com/abstract=244159.
 Robert McDonald, Fundamentals of Derivatives Markets (Upper Saddle River, NJ: Prentice Hall, 2008), 31,139.
 N. Gregory Mankiw, Macroeconomics, 7th ed. (New York: Worth, 2009), 287–321.
 Frederic Mishkin, “Is Monetary Policy Effective During Financial Crises?” (NBER, January 2009), 1–12, http://www.nber.org/papers/w14678.pdf?new_window=1.
 Federal Reserve Bank of New York, “Open Market Operations” (Federal Reserve Bank of New York, August 2007), http://www.newyorkfed.org/aboutthefed/fedpoint/fed32.html.
 Board of Governors of the Federal Reserve System, “Credit and Liquidity Programs and the Balance Sheet,” Federal Reserve, July 30, 2012, http://www.federalreserve.gov/monetarypolicy/bst_lendingdepository.htm.
 Johnathan Schwabish and Courtney Griffith, Troubled Asset Relief Program, Web, March 28, 2012, http://www.cbo.gov/sites/default/files/cbofiles/images/pubs-images/43xxx/TARP_3_27_2012.png.
 Colin Barr, “TARP Profit Claim Bugs Skeptic” (CNNMoney, August 4, 2009), http://money.cnn.com/2009/08/04/news/economy/tarp.profit.fortune/; “Bailout: Will It Work?” (CNNMoney, October 4, 2008), http://money.cnn.com/2008/10/04/news/economy/will_it_work/index.htm.
 “Financial Turmoil Timeline” (Federal Reserve Bank of New York, December 31, 2010), http://www.newyorkfed.org/research/global_economy/Crisis_Timeline.pdf.
 Michael J. Fleming and Nicholas J. Klagge, “The Federal Reserve’s Foreign Exchange Swap Lines,” Current Issues in Economics and Finance, April 2010.
 “The CBOE Volatality Index” (Chicago Board of Exchange, 2009), 1–3.
 McDonald, Fundamentals of Derivatives Markets, 39,46.
 Robert Whaley, “Understanding VIX,” November 6, 2008, 7, http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1296743.
 Cheekiat Low, “The Fear and Exuberance from Implied Volatility of S&P 100 Index Options,” The Journal of Business 77, no. 3 (July 2004): 527–546.
 “Budget of the United States Government, Fiscal Year 2013” (Office of Management and Budget, 2010), 1431–1436, http://www.whitehouse.gov/sites/default/files/omb/budget/fy2013/assets/budget.pdf.