1

DERIVATIVES, FISCAL POLICYAND FINANCIAL STABILITY

Chiara Oldani, Paolo Savona *

Luiss Guido Carli UniversitySeptember 2004

Provisional Version

Key words: fiscal policy, financial stability, derivatives andsecuritisation.JEL Classification numbers: H8, G2, G3, G19.

AbstractThe massive use of derivatives and securitisation by sovereign States for publicdebt and deficit management is a growing phenomenon in financial markets.Financial innovation can alter the stability of the public sector finance, andmodify risks effectively run. The experience of some developed and developingcountries is surveyed to look at main instruments used and aims of publicfinance. Financial stability of the public sector is analysed considering financialinnovation use. The case of Italy and its scarce disclosure of information ispresented. An IS-LM model is used to capture the effect of financialinnovation on fiscal policy, starting from Blanchard (1981). The use offinancial innovation can have positive effects over debt and deficitmanagement in the short run as far as risks are properly considered,expectations of fiscal policy are coherent with that of markets, and noexogenous shock occurs.

* Adjunct Professor and Full Professor respectively of Economic Policy at Luiss Guido CarliUniversity, Rome. Email: coldani@luiss.it savona@luiss.it.We wish to thank Mr Nardelli of theItalian Ministry of Economy for providing useful information.

2

1. Financial Stability of the Public SectorFinancial stability of the Public sector in an industrialised country depends

on many interdependent factors; the analysis shall be developed at a yearlybudget level (deficit, surplus and the structure of expenses), looking at the debtoutstanding (its evolution and reimbursement) and at fiscal rules andinternational linkages.

A simplified scheme of the public sector financial flows is:

State

G TDEPMBtB

Where G is public expenditure, T is total tax revenue, DEP is the amountof saving deposits hold by household (in Italy at Poste Italiane), MBt is themoney base created by the central bank for the Treasury1, and B is the flow ofdebt and its interest service payments (being them at fixed or floating rates).

Expenses (G) of an industrialised country are related to main duties and canvary depending on political willing; main areas of the public sector (pensionsystem, health care system, public purchases, investments and defence) can beinfluenced by international rules and events, like terrorism.

An interesting phenomenon is the international fiscal competition(especially regarding corporate taxes, a part of T); looking at Europe, Ireland isthe example, while in the US Delaware has a favourable tax system. Taxcoordination is particularly difficult with international tax evasion and elusionpossibilities for corporate taxpayers and high-income households who canchoose where to settle their headquarters. International rules about fiscalheavens have decreased tax evasion, under the political pressure coming fromthe need to know international terrorism funding, but elusion is still anunsolved problem.

Tax competition can also attract high skilled and educated workers, asshown by Giannoccolo (2003), and contribute to induce growth; Ireland andthe amazing growth of its high-technology sector is the empirical verificationof this theory.

Corporate tax level in Europe is on average around 31-32% (with adecreasing path over the last ten years) but in Ireland is 12.5%2 and in CentralEuropean Countries around 20% average. "We want fiscal competition, but wemust avoid fiscal dumping," said Gerard Schroeder, Chancellor of Germany inApril 2004, with respect to Central Europe new accessing countries and their

1 MBt is actually zero since European rules do not allow any monetisation of debt or deficit.2 In Ireland corporate tax rate is going to be higher in the future.

3

favorable fiscal rules. Newly incoming members of the European Union haveincentives not to higher their tax rate (corporate and households) to attractbusinesses from the outside, lacking many public services and having a weakbanking and financial system as a support of growth; generally speaking smallcountries do not benefit from tax cooperation because they cannot offer thesame amount of public services provided by large countries so that they engagein a price war on the tax rate.

Harmonisation of fiscal rules in Europe is very difficult, given the majorityvoting system, since interests of small countries can be in conflict with those ofbig. In particular, the rules governing the capital tax rate, value added tax andfinancial market investments tax rate are under analysis since the ‘90s, but notmuch has been harmonised.

Financial stability of the public sector can suffer because of international (ordomestic, if you consider Europe as “home”) fiscal competition and dumping;a decreasing growth rate of GDP, as we experienced recently, induces a loss invalue added tax, income and corporate taxes, but favourable fiscal treatmentscan induce the emigration of firms and headquarters and then exacerbate thepublic sector deficit, as experienced by Germany, France and partially by Italy.

Deposits and savings (DEP) hold by households and deposited within thepublic sector, for example trough the postal or public banking systems, are anexample of a way of financing public expenditure (G) since the State inindustrialised countries is high rated, and then gives low rate of return.Privatisation of public postal services and banks lead to an increasing cost ofinterest to be paid, since the 100% guarantee of the State disappears and therating lowers. An example is Poste Italiane, which manages the postal service andraises funds managed by the Cassa Depositi e Prestiti (CDP). In 1998 Poste Italianeand in 2003 the CDP have been privatised, and 35% of the stakes of PosteItaliane have been given to CDP by the shareholder of both, the ItalianTreasury. The listing of Poste Italiane on the Italian stock exchange in 2004 hasbeen announced and is part of the national privatisation process.

Flows of new public debt (B) is composed by new bonds and by interests tobe paid on them.

High-indebted countries can experience deficit (G-T>0), which pushesfurther the debt issuing; but bond issuing can be used to pay back interest (It)or capital on existing debt, pushing the growth of the debt into a self-fulfillingprocess. Italy has experienced all these phases and after 1992 has exploitedcurrent account surplus (G-T<0) to decrease the debt outstanding.

Bonds need to be issued any way, especially when forecasts of publicexpense and revenues are not fulfilled, and to compete with many high ratedcountries running into deficit over the last 2-3 years, like the US, Italy hasissued inflation linked bonds to attract funds3. In this way the real returnbefore tax of the bond should be guaranteed; this represents a further incentive

3 Ministero dell’Economia, 2003 and 2004.

4

not to inflate for the State. The efficiency for investors and the Public sector ofthis type of bond has been shown, but they represent a response to the slowingdemand of bonds experienced over the last 12 months for Treasury Bills,which yielded negative real rate of return.

Risk management of the public sector is complex and depends on whomand on which time horizon is made. Looking at Italy, at a yearly level, theBudget Law (Legge Finanziaria) addresses this issue, while over time we shouldlook at the Triennial Economic and Financial Programme.

The Italian debt outstanding in 2004 is about 106% of GDP, and equal to1.381.574 million euro, but expected to decrease after 2006. Corporate tax rateis supposed to decrease after 2005 to balance domestic (European) andinternational fiscal competition, and low-income households should experiencetax saving. This should not let the deficit to increase (G-T), given Europeancriteria. Privatisations and public wealth and estates better management shouldprovide resources.

[Tab. 1 about here]

Risk connected with high public spending of local authorities (G), anddemographic events (ageing of population, immigration, low birth rate) areconsidered together with other variables’ behaviour, but precise data about theamount of hedging and portfolio strategies of local and central bodies aremissing.

The public sector is always exposed to liquidity risk, since it pays on amonthly basis wages, salaries and various current expenses, while it receivesmany revenues on a less frequent basis (for example twice a year for personalincome taxes). Liquidity risk affects assets and liabilities of the State (G,T,DEP and B).

Public sector financial stability is related to cash management and loweringliquidity risk. Derivatives are useful for risk and cash management, but canexacerbate existing risks (market, credit, liquidity, counterparty, etc.), if aslowing down in demand of Treasury Bills and/or an unexpected shockhappens; derivatives can introduce a new component to the existing liquidityrisk of the Public Sector (like in the Orange County bankruptcy case).

As suggested by the Constitutional Auditors (Corte dei Conti in Italy), thedetails about the exposition to innovative instruments by the public sectorshould be known, and potential effects and risks addressed in public economicdocuments, to increase accountability and transparency.

2. Financial Innovation and Fiscal Policy: Some Countries’ ExperienceFinancial innovation influences modern fiscal policy in two different ways:

first, it helps in tax saving by taxpayers; secondly, financial innovation can beused by the State itself (centrally or locally) to lower the cost of debt, toimprove the cash and debt management, and reduce costs (OECD, 2002).

5

Instruments used by the taxpayers are derivatives, while the public sectorcan use derivatives and/or securitisation for debt management. Securitisation isa way to pool together credits and other financial assets (Assets BackedSecurity, ABS), sell them on the market to institutions, which utilisesecuritisation to finance their business. Assets are generally held by tax neutralvehicle (Special Purpose Vehicle, SPV), and it issues rated debt to fund thepurchase of these assets. Derivatives used for public debt management areswaps, FRAs and many others, depending on needs, debt structure andcharacteristics.

Looking at taxpayers, derivatives’ strategies are useful for tax timing option,i.e. postponing revenues and realising losses, as to lower the total amount ofrevenues, and then taxes to be paid; this has been shown for firms andhouseholds (see Zeng, 2004 and Salcedo, 2003 respectively) and then induce aloss in total revenues of the State.

By helping taxpayers to lower their burden, derivatives confirm their nature,i.e. are used to shift risks and satisfy needs of customers, being more efficientthan traditional financial instruments. The tax saving has been demonstrated inthe US tax system by many authors and will not be taken as given; theEuropean tax system is highly fragmented, and recently the EuropeanCommission has asked for homogenous definitions of what to tax and bywhom, but not yet on how much, leaving it to the freedom of countries. EachEuropean country has a different tax system for financial revenues and itincreases tax arbitraging, since it is theoretically possible to move from acountry to another and save. This could be confirmed looking at firms, whichprefer lower tax rate countries, like Ireland (Fondazione La Malfa, 2002).

Benefits of financial innovation’s use by Sovereign State are that it is a off-balance sheet item and increases funds available to public sector, given budgetcriteria (e.g. Maastrict or IMF); increases international transparency, sincecapital markets are under intense international scrutiny; is alternative toprivatisation, which is not always the sole solution to exploit public goods;diversifies investments and betters debt management. The understandabilityand disclosure of the strategy is very important for the market to believe in asovereign State finance.

Examples of countries actively using financial innovation to manage debtand deficit (locally and centrally) are known and we can briefly summarisethem basing on a rough distinction: developed (Austria, Denmark, Greece,Italy) and developing (Brazil, Hungary, India, Israel) (OECD, 2002).

Developed countriesAustria uses derivatives since 1981 such as swaps, for long-term

management, and FRAs, for short term, on interest and exchange rates. TheAustrian Federal Financing Agency is the external agency in charge to raisefunds and restructure portfolio; particular attention is due to credit risk andliquidity management.

Denmark uses extensively derivatives for debt management and

6

instruments are basically swaps: currency, interest rates, structured, liability,asset and portfolio. In 2001 the total principal amount outstanding was 121billion Danish krona (16 billion euro). Aim of this extensive use is to lower“long-term borrowing costs, while taking into account risks associated with thedebt”4. Strict rating requirements support derivatives purchases together withthe tendency toward plain vanilla style contract.

Greece has used securitisation as a debt management instrument and hassecuritised credits coming from lottery, air traffic, and revenue from the EU.The amount of securitised assets is however, much lower than other Europeancountries, and reached 3,745 million euro.

Italy with its 106% debt over GDP ratio in 2003 is one of the most sensiblecountry to debt management problems; securitisation is one of the instrumentsused to hedge public debt and it has been applied to the National Institute ofSocial Security (INPS), and to the Public Real Estate. Credits have beensecuritised and performance were different in the two cases, because of theirdifferent nature. Credits of National Security are financial assets which can betraded on the market, domestically and abroad, without much difficulty, whileItalian public real estates have an incredible burden of rules, limits andprivileges which let their trading more complicate and time consuming.However, both operations were successful and raise funds up to 9 billioneuro5. The use of derivative instruments by the Italian public sector will beanalysed in depth in the following paragraph.

Developing countriesBrazil has been hit by external shock over 2001 and effects last for 2002

and over. Currency depreciated by 40% in 9 months of 2001, FDI lowered andinflation increased; derivatives have been used to hedge against this adverseshock and instruments are dollar futures, interest rate futures, interest andexchange rates swaps, and forwards, for a total 170 billion dollars (December30th 2001). Monetary policy intervened to enhance liquidity in the market andincrease overnight target rate. Monetary and fiscal authority worked togetherto manage foreign exchange denominated debt and not boosting the exchangerate, issuing a dollar indexed bonds and supplying hedge to the market.

Hungary, as a country willing to access the European Union, has to controlmonetary and real variables under strict rules; public debt outstanding reachedthe value of 30 billion euro, which is small compared to other Europeancountries, and securitisation has been chosen as a debt and risk managementtechnique. A marginal role is given to these innovative instruments, sincemarket risks can be influenced.

India has 63.7% debt over GDP ratio and uses extensively financialinnovation to manage its costs. At March 2002 there were $10 billion ofderivatives transactions outstanding. Derivatives allowed are swaps and FRAs,

4 Cf. “The role of derivatives in Danish debt policy”, in OECD, 2002.5 Cf. “The role of securitisation of public assets: the Italian experience”, in OECD, 2002.

7

written on interest and exchange rates and in various forms (caps, collars). Themassive use of these instruments has growth dramatically over the ‘90s and‘00s because of increasing deficit and internationalisation of trading.Derivatives are used to manage risks, increase liquidity of markets, attractinginvestors, and providing shorter dates on markets.

Israel has debt over GDP ratio at 96% in 2001, 26% of which is foreign,and has introduced a single debt manager to enhance risk management, buildan infrastructure for advanced pricing capability, and to find an optimalbenchmark for liabilities portfolio. Derivatives are used “strategically inrestructuring the liabilities portfolio vis-à-vis the benchmark”. Instrumentsused are swaps, collars, swap-options; derivatives are chosen basically on risk-cost measures (efficient frontier) and to reshape the portfolio according to thebenchmark.

In its reducing costs trough the use of innovation, especially interest rateand currency swaps, what need to be carefully considered are risks effectivelytaken by the State. Credit, liquidity and market risks are related to derivatives’use. Many crashes of firms and banks have been caused by some form of mis-management (Barings, LTCM, and many others). Often only credit and marketrisks are considered, so that if a liquidity problem arises, it can have very badeffects. One example of public institution which did not considered liquidityrisk in its derivatives investing strategy was the Orange County (California)which bankrupted in 1994, after having realised 1,6 billion losses. As explainedby Marthinsen (2003), the loss was due to mismanagement of funds by theCounty, which was unable to consider risk effectively run. The poor controland monitoring systems were unable to look after what was happening toCounty’s funds, which realised high revenues with very aggressive (and un-hedged) financial operations, managed by Mr Citron, a self-educated blue-collar of the County. Total portfolio of Orange County amounted to 7.6 billiondollars, all excess funds of the County and of 200 other municipal entities(schools, hospitals and so on). The Federal Reserve decided in February 1994to rise interest rates to avoid the US economy overheating6; portfolio ofOrange County was made mainly by structured notes, fixed-income securitiesand inverse floating-rates notes, all interest rate sensible assets. Leverage ofportfolio was more than 2.5 (trough the use of reverse repurchase agreements).The fund manager, Mr Citron, has bet on falling interest rates in designing theportfolio structure, so that in front of an unexpected monetary and creditrestriction, at the beginning, he believed in the goodness of the strategy anddoubled all positions7. At the end of 1994 the return of investment was–38.55%8. The news that the County had a bad structured portfolio induce allmarket players to ask back funds and close all positions with the County.Liquidity restrictions (and not credit or market risks) lead to the bankrupt of 6 Treasury Bill rate moved from 3.54 to 7.14% over 1994.7 This was the same mistake made by Mr. Leeson, which lead to the bankrupt of Barings Bank.8 See Marthinsen, (2003) cap. 6 for data and details.

8

the County, which asked for Chapter 9 creditor protection.The role of derivatives in the crash is limited to the reverse repurchase

agreements (a type of forward contract). However, inverse floating-rates notesare considered as derivatives-type instruments, so that the total derivatives-related losses of the County was about 700 million dollars9. The general lessonto be learned by the Orange County crash is that safety, liquidity and high yieldare not possible to reach together; an opaque and complicate investing strategycan create much risk than costs it saves (or profits it makes).

3. The Case of Italy: Public Debt and DerivativesLocal municipalities, thanks to increasing devolution of fiscal sovereignty,

and of a fixed percentage of national income tax, must finance public services,like education or health care, and infrastructures, like roads, and transports.This has contributed to increase the fiscal pressure over taxpayers; Cities andMunicipalities have introduced new taxes on real estate10, and Regions haveintroduced a tax on firms’ revenues11. On the other hand, they had to financetheir deficit by issuing bonds. These bonds pay interest rate to holders, theyhave looked for hedging strategies against adverse interest rates’ movements.Banks and financial advisors are the economic institutions, which develop newinstruments to satisfy the needs of customers, gain a profit and move thefrontier of market. Local municipalities gain in the short run from hedgingstrategies and succeed in reducing costs, but it is not clear which is the burdenof costs that can potentially come out over a longer period of time. Let’s makean example.

Suppose the Lazio Region needs 100 and issues a 5-year bond paying theEuribor rate + 100 basis points annually. Today, the Euribor is less than 3%,so that the total payment is less than 4%. But expectations on interest rates saythat they are going to rise over the next 5 years, so better to hedge against thatevent not to increase the interest rate costs over time. The Bank offers a 5-yearinterest rate swap; expectations about how much the interest rates can rise areof central importance, together with the time period considered for thecontract. If expectations of the Region is that the interest rate can rise, but notmuch (not a ‘80s level, when interest rates where above 8%), and particularlythat interest rates will not be above 5%, they could be offered by the Bank thefollowing type of contract:

9 Marthinsen, by confronting different replication strategies of the portfolio, concludes thatderivatives’ role in the bankrupt can be considered as much smaller (about 330 million dollarsover the total 1.6 billion) and poses some doubts about the effective liquidity crisis of theCounty.10 The Imposta Comunale sugli Immobili (ICI).11 The Imposta sui Redditi delle Attività Produttive (IRAP).

9

Lazio Region pays flow of funds Bank paysEuribor - 100 Euribor – 100 if Euribor is less than 5%Euribor + 300 Euribor +100 if Euribor is greater than 5%

Holders of bondsEuribor +100

If the interest rate, the Euribor, is below 5% the Lazio Region gains becauseit pays the Euribor – 100 to the Bank which pays Euribor + 100 to the bond’sholder; the saving for the Region is 200 basis points of the bond nominal valueon an annual basis. If the Euribor is above 5%, the total amount paid is at least8%, since it pays Euribor + 300 b.p. to the banks. The costs is much higherthat the initial.

One key problem is the expectation formation process, and over whichperiod of times the “bet” is made. If the electoral mandate is less than 5 years,let’s say 2 years, and expectations over two years are that interest rates will notrise above 5%, there could be an incentive to choose the above written swapcontract, saving more and using it to be re-elected and then swap the aboveswap contract if there is the need (i.e. if the interest rate goes above 5%).Expectations play a key role in the choice of the hedging strategy together withincentives.

The Corte dei Conti, the administrative controller of Italian public accountingand practises, has stated that swap can be used to manage the lower resourcesavailable from the centre to the periphery of the State; specifically, the interestrate swap is designed to exchange interest rates paid on bonds issued ondomestic and international markets; this practice is allowed since 2001 by theState budget law. Tuscany Region, Sicily Region, Provinces of Varese andPavia, Cities of La Spezia, Reggio Emilia, Udine and Venice have used swapsadvised by J. P. Morgan Chase and other prestigious banks. The Lazio Regionhas opened an Office devoted to help municipalities in the Region to developthe best hedging strategy and saving costs. The public sector has to update itsknowledge and exploit new financing means and instruments, and their well-known advantages.

In a speech at the Italian Senate of the Republic in March 2004, the GeneralDirector of the Finance Ministry, and actually deputy Minister, DomenicoSiniscalco (2004a), had explained the way the public sector uses derivatives,and guarantees that risks are properly addressed. Some local municipalitieshave been a little bit too aggressive and maybe not much cautious, but soundmonitoring and control are guaranteed.

Italian public debt is composed at 54% by long-term fixed rate bonds (BuoniPoliennali del Tesoro, BPT), and at 31% by short and long term floating ratebonds (Buoni Ordinari del Tesoro, BOT and Certificati di Credito del Tesoro, CCT).The share of floating rate bonds is around 25-30%, and the debt’s costs havebeen lowered in the last 10 years, from 14.05% to 3.8%, thanks also to

10

lowering European interest rates. Average life of debt has increased, togetherwith the duration over the last decade, and reached 5.9 years. Issuancetechniques of the Italian Governement are auctions, syndications, exchangeoffers and various combinations of the latter (Ministero dell’Economia, 2003).Using syndicated deals long-term bonds, foreign currency denominated bondsand innovative instruments are placed.

A bond exchange program has been allowed since 2002 to manage risksmore efficiently, to obtain a smoother debt redemption profile, and to enhancethe liquidity of the secondary market (Ministero dell’Economia, 2004a). Debtmanagement policy in 2003 has been oriented to new benchmarks on the longpart of the yield curve, has introduced new inflation-linked bonds, and haslowered the amount of floating rate bonds (Ministero dell’Economia, 2004b).

Debt management in 2004 will follow guidelines for 2003, and issue at least8 billion US$ denominated bonds, and 2-2.5 billion in other markets for a total10 billion euros (Ministero dell’Economia, 2004b).

Recently, the Minister of Economy, Siniscalco (2004b), has given somenumbers about the activity of local public authorities (Regions, Cities,Provinces, and Municipalities) on the use of derivatives and their purposes.The Italian Treasury admits the lack of data about this activity, which hasreached 856 million euro notional amount total from February to July 2004only, and the dynamic evolution of financial markets is an objective obstacleagainst a clear picture of instruments, counterparts and markets involved. 35%of Italian Regions, 31% of Municipalities and 28% of Provinces engage inswap activities, but many local authorities have not provided complete data.What is remarkable is that cost saving is not one of the main reasons for theItalian Governement to use derivatives, since it is less than 1% of notionalvalue, compared to open market operation (Table 2).

The Minister sets which types of derivatives can be traded by publicauthorities, and these are: plain vanilla interest rate swap, interest rate cap,interest rate collar, and forward rate agreements (FRAs). All derivatives shouldbe “plain vanilla” style, i.e. no derivatives on derivatives, no exotic or structuralinstruments linked to any principal. In particular, knock-in swaps are forbidden(i.e. if the Euribor reaches a pre-defined - high - level, the authority pays twicethe Euribor to the counterpart). Positions are limited too, in particular longand short position can be taken on swaps and FRAs, but only long positionson cap and collar are allowed. In Table 2 there are some qualitativeinformation provided by the Ministry of Economy on request, where optionsare, however, still present.

The problem of restructuring existing derivatives liabilities, which are notallowed any more (like options), is solved by saying that the relative costscannot be shifted to future budget years (when derivatives effectively expire).Risks should be bear (and paid) by who raised them. The problem of theauthority who should look after derivatives’ trading has been solved, since theTreasury is the only able to allow for derivatives purchases by local authorities.

11

The database of derivatives activity by local authorities is not accessible, andthe Treasury admits having some troubles in collecting complete data aboutderivatives from Regions (Siniscalco, 2004, p. 9). No punishment or fine isruled for those providing false (or even no) data.

The picture of derivatives’ use by the Public sector is not complete at themoment, since public data are not provided. They should be given accordingalso to European savings protection principles. Every financial operation hasimmediate effects but induces some forms of risk, which can show up in thefuture. This dilemma is as old as the State itself: the incentive to cheat, gain andbe re-elected can overcome the potential costs of the worse scenario.

Many other countries, industrialised and developing, engage in derivativesactivity to hedge on domestic and international markets12. Plain vanilla swapsand options are widely used and back office procedures and control are crucialnot to raise other risks.

Denmark and Australia, for example, use interest rate and currency swaps tolower long-term borrowing costs. Only high rated counterpart (A3, AA) can beinvolved in such operations; preference toward plain vanilla type of contracts isgiven, together with and the inclusion of cross-default clauses, which helpdecreasing credit risks. Other developing countries, having high external debtlike Brazil, use derivatives to hedge and lower the cost of debt. Monetarypolicy cooperation plays a central role in this last case, since monetary andeconomic conditions are much complicate, and the rating of these countries isnot very high.

4. Implications of Derivatives’ Use by Fiscal AuthoritiesThe use of financial innovation and derivatives by public sector has certain

cost savings’ effects, and brings benefits to national and international financialmarkets, increasing liquidity and efficiency of public debt management. Thegrowth rate of these markets and contracts is justified by their efficiency in costsaving and pricing systems, and by their liquidity and flexibility; the publicsector, as a player of financial markets, benefits from financial innovationimprovements.

However, financial innovation directly increases existing risks (market,credit, liquidity and counterparty) over a pre-determined time length, and thencould act in the opposite direction of financial stability. OTC derivatives canindirectly induce different forms of risk thanks to the opacity of trading, lowtransparency of settlements systems, and scarce accounting and registrationprinciples; moreover, the interaction between central and local publicauthorities using financial innovation can alter financial equilibrium and modifythe allocation process of resources from the centre to the periphery.

12 For details see OECD, 2002.

12

A public sector characterised by a heavy burden of debt outstanding shouldexhibit low risk loving behaviour, not to increase existing risks; using financialinnovation for debt and cash management has positive effects, but in case ofshocks can exacerbate risks, giving raise to a different cost-return of portfolio.

An active use of derivatives, for hedging and speculation, can be consideredas an indirect source of financial instability and influence the investment-savingrelationship of the public sector. The I-S relationship is dependent onsensitivity to invest and save, measured by the slope of the IS curve (steeper orflatter); the sensitivity influences the elasticity of the curve with respect toincome (Y), a part of which is made by investment, and the interest rate (r), theprice of investment (and savings). Financial innovation, whose use is based onexpectations, by influencing the ability of the State to borrow on market andinducing new risks, trough leverage effects, increases the instability of the I-Srelationship (its slope can change if adverse shocks take place).

The indirect effects of derivatives are also to be linked with the tax timingoptions of tax-payers (firms and households), enhancing liquidity risks for theState. At a macroeconomic level the tax timing option can influence therelationship between investments and savings, since the private sector can havemore resources at disposal and exhibit more risk loving behaviour, increasingthe IS curve slope.

A very active use of derivatives by private and public sectors can be asource of “real” instability; derivatives affect financial markets, increasinginstability in case a shock occurs, but a comprehensive analysis of monetaryaspects is left to a separate research.

We can use these intuitions to modify existing macro model and look at theeffects, positive or not, of the use of financial innovation (securitisation,derivatives) by the public sector and analyse policy behaviour, given previouscountry experience.

5. A Macro Model Considering ExpectationsBlanchard (1981) extended the traditional IS-LM model to consider the role

of expectations, of asset prices and their interaction with output. We willproceed with summarising Blanchard model in this paragraph, and in followingsections will proceed with extensions.

Main hypotheses are that “the economy is closed, the physical capital stockis constant, there are one good and four marketable assets, which are shares ofphysical capital, there are private short term and long term bonds, issued andheld by individuals, and outside money”13. There are three main spendingdeterminants: “shares of the stock market, which being short term, influencesconsumption, and influences investments, by determining capital value, trough

13 See O. Blanchard (1981) page 132.

13

its replacement cost” (Tobin’s Q). Output depends on stock market and fiscalpolicy.

Real spending (d) is composed by the stock market value (q), real income (y),and influenced by fiscal policy (g).

1.

€

d = aq + βy + g;a > 0;0 < β ≤1

An interesting consequence of equation 1 is that, following an increase inaggregate demand, inventories are decumulated first, and then production isincreased. Output adjusts to spending over time, so that its growth rate can beexpressed as:

2.

€

y•

=σ(d − y) =σ (aq + g − by)σ > 0 b ≡1−β > 0

Spending equals production, but actual spending adjusts slowly to desiredspending d.

The asset market is in equilibrium if there is no arbitrage on short termreturn of the three assets; the portfolio balance is characterised by LM curve:

3.

€

i = cy − h(m − p)c > 0;h > 0

r* ≡ i − p•

*where i is short term nominal interest rate, y is real income, m is log of nominalmoney, p is log of price level, r is real short term rate, r* is the expected level of

the rate, and

€

p•

*is expected rate of inflation. By imposing no arbitrage betweenlong and short term bonds, and considering R as long term interest rate we getthat:

4

€

r* = R − R•

* /RBy considering the no arbitrage condition between short-term bonds and

shares, if q is real value of stock market, the expected real rate of return onholding shares is

€

q•

* /q + π /q, where π is real profit, and

€

π =α0 +α1y α1 ≥ 0Following Tobin, the link between financial and real markets is the value of

stock market, q. The no arbitrage condition between shares and bonds on longand short term leads to:

5.

€

q•

*q

+α0 +α1y

q= r *

To close the model let’s suppose expectations are rational, so that we missonly the equation of price level behaviour.

14

First Blanchard simplifies the system by hypothesising fixed prices, so thatactual and expected inflation is zero; nominal and real variables coincide.

At the steady state, the system of equations simplifies to:

6.

€

y•

=σ(aq − by + g)r = cy − h(m − p)

q•

*q

+α0 +α1y

q= r

r = R − R•

RExpected rate of interest (r*) has been replaced by r, and last equation is the

term structure.Blanchard (1981) looked at the effects of monetary and fiscal expansions

under fixed price hypothesis and at monetary expansion under flexible price;the effects of a discrete change in current or anticipated policy is a discretechange in the stock market due to the change in the anticipated sequence ofprofits and interest rates. This, in turn, affects spending and output over time.Output and the stock market change because of a policy movement. Theannouncement effect plays a central role in the effectiveness of policymanoeuvre, since it can change anticipated profits and discount rate, leading toan effective change of the stock market14. The Author underlines a perverseeffect of fiscal policy announcement, since it can decrease output, because ofcrowding out effect, before the policy is even implemented; this is mainly dueto the rational expectations hypothesis.

6. A Modified Model to Consider Financial Innovation’s Use in FiscalPolicy

In the following sections we will try to modify Blanchard model to considerexplicitly financial innovation as a tool of fiscal policy; Blanchard model isflexible enough to model a modern fiscal policy behaviour, consideringmonetary policy as given and not dependent from fiscal policy, and a portfolioapproach. We consider as given the separation between the two policyauthorities and do not alter any characteristic of monetary policy with respectto the Blanchard original model. We are conscious that financial innovationalters financial market’s behaviour as well (i.e. LM curve), and this should bedirectly modelled. This issue has been addressed by Savona (2004) and will befurther developed in a separate research project.

The introduction of financial innovation use into fiscal policy behaviour (g)changes slightly the specification of the model, and its main implications. The

14 See O. Blanchard (1981) page 141.

15

distinction between short and long-term adjustment will be given. An activeuse of financial innovation is particularly interesting for those countriescharacterised by high deficit and debt, either domestic or foreign currencydenominated; financial innovation can be particularly useful for cashmanagement and hedging, to lower costs and bettering portfolio balancing.Our analysis will be focused on heavily indebted countries, or running highdeficit like Italy, the U.S., and Brazil.

The index of fiscal policy, g, contrary to Blanchard model, where it wasconsidered as full employment deficit target, here has to be considered as thedebt-deficit level targets. Decreasing debt and deficit targets are assimilated to arestrictive fiscal policy, and this comes with a lowering public expenditure,decreasing national spending (d). Fiscal policy uses financial innovation tohedge and, basing on some expectations on future interest rates level, fixes theamount of interest to be paid on bonds, in between a certain corridor of rates.To introduce financial innovation we need to consider expectations15.

We can introduce expectations in the model, by assuming that the noarbitrage condition considers risk premium (χ).

7.

€

R − R•

R= r + χ = i − p*+χ

Expectations are the most important ingredient in a portfolio composed ofbonds, money and financial innovation; financial innovation’s use changes theclassical IS curve setting, since the target of debt and deficit, being dependenton interest rate but not directly on output, can be considered as very elastic tointerest rate, but not (much) to income in the short run. The dynamic of debtin each period of time is given by

8.

€

Bt = It + Bt−1 + DEFt ; DEFt =Gt −Tt =Gt − tYt ; It = rBt−1Elasticity of debt and deficit to income (Y) and the interest rate (r) depend

on the structure of the economy; if the debt outstanding is high, deficit shoulddecrease over time (like for European countries where surplus, G-T<0, is usedto pay back existing debt, Bt-1) and the sensitivity of fiscal policy to interest ratecan be considered to be higher than that to income. If public debt outstanding(B) is not high, the two elasticity become relevant for fiscal policy, and incomeplays an explicit role for fiscal policy. The IS curve has still income in itsspecification, but fiscal policy spending, g, is under control only in between acorridor of rates (

€

r* ≤ r ≤ r **), where financial market and innovationcontributes to control the dynamic evolution of debt and deficit.

9.

€

Bt = (1+ r)Bt−1 +Gt − tYtWe can synthesise this relationship of fiscal policy over interest rate and

income elasticity, depending on debt outstanding, as:

10.

€

g =ηr +ψyη >ψ if B is high

15 See O. Blanchard (1989), page 532.

16

We can imagine that the sensitivity of fiscal policy, represented by the IScurve, to income (ψ) between pre-determined interest rates, high (r**) and low(r*), is very low or even constant, since the target of costs of debt and deficit isprimary and achieved trough the use of financial innovation16. In this corridorthe sensitivity of fiscal policy to interest rates (η) is greater that that to income.Derivatives are used to lower It, the cost of debt and control Gt, over someexpectations on rates, as shown in the previous survey of countries’experiences. In this corridor we can say that income is no longer a primarytarget of fiscal policy since the target of debt-deficit cannot be achieved ifanother is followed17. Financial innovation is used to settle the cost of debt (ordeficit) g basing on some expectations over interest rates, whose level is settledby the market, and income (or unemployment if you prefer) becomes asecondary policy target. We have to underline that this trade-off betweenincome and debt-deficit targets of fiscal policy is true in the short run, sincelong run equilibrium of the model cannot depend on financial innovation,which bets over short-term rates and is based on expectations.

Starting from public spending (d) we can state that it is a function of income(y) in the short run, of debt-deficit target (g), and of stock market value (q),where the debt is managed. We can rearrange the short-term model as:

11.

€

d = aq + g + βy IS curvea > 0 β > 0Iff r* ≤ r ≤ r ** Interest rate corridorr = cy − h(m − p) LM curve

q•

*q

+α0 +α1y

q= r No - arbitrage condition

r + χ = R − R•

R Term structure

Fiscal policy index is very sensible to interest rates, so that if

€

r* ≤ r ≤ r **(expectations of the State over interest rates are satisfied) the dynamic of debtand deficit is under control, and financial innovation contributes to lower thecost of debt and public spending.

Short-term interest rates are settled through the interaction between the LMcurve, the no arbitrage condition and the term structure, so that the marketsettles interest rates, and the fiscal authority has no power to influence them.Since interest rates represent the cost of debt, financial innovation is used tosettle a pre-defined cost of debt over constant expectation on long-term rate

16 E.g. swaps or forward contracts with which a bet over interest rates is possible andcontribute to save costs.17 This is the same idea of the unholy trinity for monetary policy and the exchange rate. Herethe ingredients of the trinity are high debt-deficit outstanding, autonomous monetary policyand output target. The third target is not achievable by fiscal authority given the other two.

17

(R), and certain risk premium (χ). Fiscal policy uses derivatives to control g inthe spending function.

Into the corridor of rates,

€

r* ≤ r ≤ r **, the equilibrium between the IS, theLM and market is such that all targets are satisfied: fiscal policy reaches thedesired levels of debt and deficit, monetary policy controls money (or prices),the market settles the interest rate and income (y) is in equilibrium.

18

Outside the corridor of rates (

€

r p r*;r f r **) the IS curve reaches anequilibrium which is associated with either higher or lower debt-deficit (area 1or 3 in graph 1), if the LM curve is not moving18.

Financial innovation with constant risk aversion, χ, and no exogenousshock can be effective to control short-term costs of debt-deficit. Fixed pricelet the story to be the simplest, since the equilibrium is the desired.

If a shock occurs, changing expectations over long-term rates (R), and riskaversion (χ ), given that the no arbitrage condition is binding,

€

q•

*q

+α0 +α1y

q= r , the relevant equilibrium can lie outside the corridor, and

the system is that described by Blanchard, but with an uncontrolled debt-deficit dynamic.

Long run solution of the model (11)19 can be found by imposing thatincome equals spending (y=d) fiscal policy controls

€

g , the market expects r,and monetary policy controls

€

m ; prices are fixed (

€

p) and we can compute riskversion (χ). The system solves finally:

9.

€

y =abq +

1bg

q =πr

=α0 +α1y

cy − h(m − p)R − χ = cy − h(m − p)

The long run solution is such that output depends on fiscal policy and thestock market; the stock market is the ratio between steady state profit andinterest rate20. The two curves have the traditional shapes but come frommodified hypothesis and behaviour.

The target of debt and deficit can be reached in the short run by means ofderivatives and securitisation; in the long run risk premium and exogenousshock let the game much difficult to play since the equilibrium is set by theinteraction between market and policies.

Our first conclusion is to underline the very positive role of financialinnovation in matching short-term targets of debt-deficit (

€

g), givenautonomous monetary policy (

€

m ), financial markets setting r, fixed prices (

€

p ),but losing control over output target (y) by fiscal authority (i.e. sensitivity toincome target is lower if debt outstanding is high).

18 A further development of the model would be to allow the LM curve to react and move; thiswill be done in a separate research project.

19 Hp:

€

R•

= p•

= q•

= q•

* = 0 and solve for y, q and r, given exogenous variables.20 O. Blanchard (1981) page 134.

19

This theoretic result seems to be confirmed by the behaviour of highindebted countries, which are involved in many OTC derivatives transactionsand run high risk in exchange of lowering debt-deficit dynamic over the lastdecade.

6.1 Shocks to the model and the effectiveness of policyConsidering fixed price level, if a shock occurs and changes risk aversion of

the public sector, χ, the term structure and the equilibrium rate change,influencing the slope if the curves first, and the final equilibrium of the systemafter. If risk aversion increases, so that the public sector accepts less risk (andlower return) the short term rate r lowers, influencing capital market value. Ifthe rate falls below the lower level accepted by the public sector (

€

r ≤ r *), theIS becomes elastic with respect to market interest rates since derivativesbecome “out of the money”. The Tobin’s Q is the link between real andfinancial markets, and if market rates are lower, capital market value lowers andthe equilibrium is at lower level of all variables. Fiscal policy could have notreached its target of debt-deficit, being at a lower level.

If a shock occurs and lowers expectations over capital market value (q*) thefinal effect is the same as described above.

Another shock can be considered a change in g, for example in a climate ofelections. If the target of fiscal policy, g, changes, e.g. increases, so that we aimat reaching a lower debt-deficit target, the effect is that of a restrictive fiscalpolicy, moving to the lower bound of rates (

€

r→ r *). This can be managedusing more financial innovation thus decreasing risk aversion (χ) accepted bythe public sector.

If a shock to financial market changes expected long-term rate (R) the termstructure changes, and rational expectations incorporate this in short-term ratesand the capital market value increases. If the interest rate falls above thehighest accepted by fiscal authority,

€

r f r **, derivatives become “out of themoney” and the final equilibrium is expansionary on output but “out of themoney” for debt management (area 1 in graph 1).

Generally speaking, unless an un-anticipated shock occurs to the economy,derivatives are very good instruments to reach desired target levels of debt anddeficit, lowering the sensitivity over income level; if a shock occurs, derivativescan exacerbate its effects, and alter financial stability of public sector, bychanging its liquidity risk and lead to any equilibrium with high undesireddeficit.

An expansionary monetary policy, moving the LM curve up to the right, canhave different effects if the final equilibrium is inside or outside the corridor ofrate (

€

r* ≤ r ≤ r **); the better solution would be to get a level of rate inside thecorridor, so that monetary and fiscal targets are reached at the same time. If theun-cooperative monetary policy manoeuvre leads to reach a rate outside thecorridor,

€

r f r **, the fiscal is in contrast with monetary policy and marketexpectations over r influence the equilibrium.

20

Our second conclusion is that fiscal policy can be considered as completelyeffective over its target of debt and deficit, without disturbing real spendingand income, if the interest rate settled by the interaction with the market is atthe desired level; in this way expectations are satisfied and no contrast betweenmonetary and fiscal policy, and the market arises. The focus then has to be putover the correct level of the rate to be expected by fiscal policy, to be coherentwith the market and monetary authority. The burden of risk implied in the useof financial innovation has to be properly considered, since can modifyfinancial stability of the public sector.

In short survey of countries using derivatives to manage cash and debt,Brazil is one paying attention to this interaction, shows low risk lovingbehaviour, and monetary policy coordination, so that the market support theinvestment and hedging strategy. Italy has shown a fragmented behaviour,since, centrally and locally, a scarce flow of information and coordination isprovided; risk loving should be low since the national burden of debt is veryhigh, but provisional data about the use of financial innovation seems reveal anaggressive behaviour. We suppose that the Italian Government has someinterest rates expectations, and up to now these have been coherent withmarkets rates. In official document we did not find any concern or explanationregarding liquidity risk of the State, which can be altered by financialinnovation’s use, or regarding adverse shock effects.

21

7. Concluding remarksWe have looked at the use of derivatives by fiscal authority and observed

that the necessary attention has not yet been paid to the link between policytargets and financial innovation’s use. Political debate and traditional economicanalysis have not focused on the effects on financial stability of public sectorusing and facing innovations; the use of derivatives is mainly, but not only,devoted to cost saving and hedging debt. Financial stability of the public sectoris strictly related to its liquidity risk, which needs a special attention,

A simple IS-LM model has been used to develop the analysis starting fromBlanchard model (1981), which introduced expectations and capital marketvalue into the traditional IS-LM framework; the author analysed theeffectiveness of anticipated and un-anticipated monetary and fiscal policymanoeuvres under rational expectations’ hypothesis, fixed and flexible prices,and effects on capital market value.

Introducing derivatives into the IS curve as debt and deficit managementtool, we reach the following conclusions; first is to underline the very positiverole of financial innovation in matching short-term targets of debt-deficit (

€

g),given autonomous monetary policy (

€

m ), financial markets setting r, fixedprices (

€

p ), but losing control over output target (y) by fiscal authority (i.e.elasticity to income is lower that that to interest rate if debt outstanding ishigh).

Our second conclusion is that fiscal policy can be considered as completelyeffective over its target of debt and deficit, without disturbing real spendingand income, if the interest rate settled by the interaction with the market is atthe desired level; in this way expectations are satisfied and no contrast betweenmonetary and fiscal policy, and the market arises. The focus then has to be putover the correct level of the rate to be expected by fiscal policy, to be coherentwith the market and monetary authority. The burden of risk implied in the useof financial innovation has to be properly considered, since can modifyfinancial stability of the public sector; the burden of risks can exacerbatenegative effects over interest rates rendering derivatives “out of the money”,and modifying debt and deficit dynamic.

With that respect, the example of the Italian Government is remarkable,since the dynamic of debt and deficit is managed trough the use of financialinnovation, centrally and locally, and is effective in the short run; on the longrun the burden of risk is not known, and only recently the Ministry ofEconomy, Siniscalco, has asked not to increase future risks by means ofinnovation and asked for restructuring portfolio of local authorities. Dataabout the future burden are not known.

The cooperation between fiscal and monetary authority, like that developedby Brazil, can lead to a better equilibrium (inside the corridor of rates), but thisis not new to economic theory.

22

8. BibliographyO. J. Blanchard (1981), Output, the Stock Market, and Interest Rate,

American Economic Review, vol. 71, n. 1, pp. 132-143.Fondazione U. La Malfa (2002), Rapporto sull’Unione Monetaria Europea

(2002), a cura di Enzo Grilli, Giorgio La Malfa, Leonardo Melosi, Laura IlariaNeri, Chiara Oldani, Mauro Piermarini e Paolo Savona, Roma,www.fondazionelamalfa.org

P. Giannoccolo (2003), Migration, brain drain and fiscal competition, WorkingPaper n. 462, Dipartimento di Economia, Università di Bologna, Bologna.

J. Marthinsen (2003), Risk takers: uses and abuses of financial derivatives, Pearson,Boston, MA.

Ministero dell’Economia e delle Finanze (2003), Government Bond IssuanceTechniques: an Overview Based on the Italian Experience, OECD Global Forum,November 28, Rome.

Ministero dell’Economia e delle Finanze (2004a), Guidelines for Public DebtManagement for 2003-2004, Dipartimento del Tesoro, Roma, www.tesoro.it .

Ministero dell’Economia e delle Finanze (2004b), Guidelines for Public DebtManagement for 2004-2005, Dipartimento del Tesoro, Roma, www.tesoro.it .

OECD (2002), 12° Workshop on Government Securities Markets and Public DebtManagement in Emerging Markets, Ministero dell’Economia e delle Finanze,Roma.

OECD (2004) Long-term Budgetary Implications of Tax-Favoured Retirement Plans,Working Paper n. 393, Economic Dept, Washington D.C..

C. Oldani (2004), I derivati finanziari: dalla Bibbia alla Enron, Franco Angeli,Milano.

M. Olson (1971), The Logic of Collective Action, Harvard University Press.H. S. Rosen (2003), Scienza delle finanze, McGraw-Hill, Milano.Y. Salcedo (2003) A time for trade, a time for taxes, Futures, April, pp. 62-

64.P. Savona (2004) Derivatives, Money and Real Income, forthcoming on

Research in Banking and Finance.P. Savona (2003) The Finance of Derivatives, Istituto Treccani.P. Savona (1975), “Inflazione, moneta e bilancia dei pagamenti. Effetti

strutturali delle politiche di breve periodo”, Rivista di Politica Economica, vol. V.D. Siniscalco, M. Cannata (2004a), Effetti e tecniche di controllo dei flussi di

finanza pubblica in ordine all’andamento del debito con particolare riferimento allacomponente non statale, Audizione alla Commissione Programmazione Economicae Bilancio, Senato della Repubblica, 24 marzo, Roma.

D. Siniscalco, M. Cannata (2004b), Effetti e tecniche di controllo dei flussi difinanza pubblica in ordine all’andamento del debito con particolare riferimento allacomponente non statale, Audizione alla Commissione Programmazione Economicae Bilancio, Senato della Repubblica, 24 luglio, Roma.

The Economist (2002), Demography in America and Europe: a tale of two bellies,Aug 22nd, London.

23

T. Zeng (2003), Tax planning using derivatives instruments and firm marketvaluation under clean market valuation, PhD dissertation Queen’s University2001, Doctoral Research in Taxation, Journal of the American Taxation Association,Spring, Vol. 25, Issue 1, pp. 142-3.

T. Zeng (2004), Tax timing options: firms use derivatives to save taxes, mimeo,Wilfried Laurier University.

9. Tables

2003 2004 Forecast

2005 Forecast

2006 Forecast

2007 Forecast

2008 Forecast

REVENUESIncome Tax 177370 181507 188000 197500 205600 216400Consumption Tax 188522 196630 203500 211600 219200 227300Tax on Capital 20204 7548 350 300 250 200Total taxes 386096 385685 391850 409400 425050 443900Social contributions 171028 178401 184000 190200 196300 203500Other current revenues 41345 42588 43000 44750 47400 48000Capital Revenues 4294 7367 5000 5000 4700 4900Total revenues 602763 614041 623850 649350 673450 700300Fiscal Pressure 42.8 41.8 40.8 40.8 40.5 40.4EXPENSESWages and Salaries 143606 152714 155800 159400 162850 167000Consumption 102280 103742 109800 114800 119900 125000Pensions 185231 193046 200500 209400 217450 227280Other Social Payments 38979 41554 41700 43300 44300 45720Other Current Expenses 42113 44928 47800 49350 49700 50400Total Current Expenses 512209 535984 555600 576250 594200 615400

Interest Payments 69291 71702 74000 79700 87100 94300

Total Current Expenses Including Interests

581500 607686 629600 655950 681300 709700

of which Health Care Exp. 81324 89650 92434 95644 99676 103856Capital Expenses 53095 47928 56900 57230 56850 55200Total Expenses Excluding Interest 565304 583912 612500 633480 651050 670600

Total Final Expenses 634595 655614 686500 713180 738150 764900Primary Current Balance* 37459 32129 11350 15870 22400 29700% 2.9 2.4 0.8 1.1 1.5 1.9Current Balance -3235 -8561 -11100 -11900 -12800 -14500% -0,2 -0,6 -0,8 -0,8 -0,8 -0,9Net Indebtness of Public Sector* -31832 -39574 -62650 -63830 -64700 -64600

% -2,4 -2,9 -4,4 -4,3 -4,2 -4Public Sector Net 42681 62000 83000 87900 85000 96000% 3.3 4.6 5.9 6 5.5 6GDP 1300926 1350128 1409769 1469830 1533966 1600505(*) In 2004 numbers are affected by 2000 million manouvre to be ruled.Source: DPEF 2005-2008, July 2004.

Table 1 Forecats of Italian Public Sector Revenues and Expenses(Million of euro)

24

Tab. 2 The Italian Public Sector: Derivatives Activity September 1th, 2004Types of contracts YES NOSwap XOption XFRAs XRepurchase agreement XForex agreement XFuture XOther XCharacteristics of ContractsUnderlyingInterest Rate XExchange Rate XCreditCommodityOtherMarket StructurePlain vanilla XStructured XExoticBarrierknock inOtherTime lenghtOvernightOne weekOne month XSix Months XLess than a year XMore than a year XMore than two years XMore than five years XMore than ten years XType of CounterpartsSovereign StatesItalian banks and financial intermediaries XForeign banks and intermediaries XItalian firmsForeign firmsOtherRating of counterpartsHigh rated XLow ratedNo ratingRating is not necessaryCosts of operations: intermediation fee(as a percentage of notional value)Less than1% XLess than 5%More than 5% but less than 10%Less than 15%More than 15%Other costs(as a percentage of notional value)Less than1% XLess than 5%More than 5% but less than 10%Less than 15%More than 15%Costs saving with respect to open market operation(as a percentage of notional value)Less than1% XLess than 5%More than 5% but less than 10%Less than 15%More than 15%