ࡱ> `bjbjss4 ```"000 BKBKBK8zKLLZqLNNNNNNNQoSoSoSoSoSoSo$Drhtvwo0TNNTTwo00NNHqfffT 0N0NQofTQoff:k,00mlNL HBKal -o$^q0qlR"ve*"vml"v0mlNbPftQPRNNNwowofNNNqTTTTZZZD1ZZZ1D8,000 THE PARADOX OF KUNA SAVINGS: WHY CROATIAN SAVE IN FOREIGN CURRENCY SUMMARY It has long been a fact Croatians save in foreign currency instead of domestic currency. This paper investigates why Croatians when faced with a stabile economy and stable currency still have major prejudice towards their own currency. The aversion towards local currency goes so far households are willing to accept negative real rates of return when saving in foreign currency. What is the reasons and rational behind this choice? The authors develop a dynamic model of households behavior and the use a Bayesian framework in order to create an objective and subjective probability distribution for making economic decisions. The empirical results also show there might be a threshold of attention where the households would move from savings in foreign currency to savings in local currency if the returns in local currency were high enough. Key words: household savings, optimal choice, dynamic programming, Bayesian choice JEL: D31, C61, C11, Gregurek Miroslav Privredna Banka Zagreb Ra koga 6 10000 Zagreb  HYPERLINK "mailto:neven.vidakovic@pbz.hr" mgregurek2@pbz.hr Neven Vidakovic Privredna Banka Zagreb Ra koga 6 10000 Zagreb  HYPERLINK "mailto:neven.vidakovic@pbz.hr" neven.vidakovic@pbz.hr 1. Uvod Od konstituiranja Hrvatske dr~ave, preferira se devizna atednja. Stanovniatvo u Hrvatskoj ima jaku averziju prema uvanju uateevina u domaoj valuti. Averzija ide tako daleko, da su ljudi skloniji prihvatiti negativnu realnu stopu povrata na svoje uateevine u devizama, nego uvati uateevine u domaoj valuti i na taj na in ostvarivati pozitivnu stopu povrata. Ovaj rad istra~uje i pokuaava objasniti taj paradoks u ekonomskom ponaaanju. U proalosti je izbor kao valuta za deviznu atednju u Hrvatskoj bila Njema ka marka, a implementacijom Eura, atanovniatvo izabire tu valutu za uvanje svoje uateevine. Krajem 2007. godine, atednja stanovniatva u Hrvatskoj iznosi 83,2 milijarde HRK u inozemnim valutama, a 23,4 milijarde u HRK u domaoj valuti. Razvidno iz ovog odnosa atednje u inozemnim i domaoj valuti, odnos je 3 : 1. Namee se prirodno pitanje: ato pokree stanovniatvo da ne atede u vlastitoj valuti, a atede u devizama? Kako je vidljivo iz podataka (Bilten HNB), to ponaaanje nije niata novo, a kako se vraamo u povijesne vremenske serije Biltena HNB, preferiranje atednje u devizama egzistira u svim podacima. Da bi u potpunosti objasnili taj fenomen, moramo koristiti dva alternativna pristupa. Prvi pristup je psiholoaki a drugi je ekonomski vremenski problem. U psiholoakom pristupu uzima se injenica da je Hrvatska ekonomija relativno nova ekonomija, koja je joa u procesu tranzicije od socijalisti ko upravlja ke, dogovorne ekonomije prema tr~ianoj ekonomiji. Smjer u kom Hrvatska mo~e ubla~iti ili potaknuti uspjeh na tom polju je razvidno u radu Gregureka i Vidakovia (2007), gdje je tranzicija prezentirana kroz teoriju igara i s analizom krajnjih rezultata. Od neovisnosti Hrvatske a nakon zavraetka rata, glavnina ekonomije se obnovila. Jednom kada je rat zbio zavraen, nova kapitalisti ka ekonomija je zapo ela s ustrojavanjem. Uspjeh nove ekonomije kreiran nakon socijalisti kog ekonomskog ureenja nije joa determiniran, dok se god tranzicijski proces ne zavrai u pridru~ivanju Europskoj Uniji i EMU. Ti elementi nesigurnosti izazivaju strah i nepovjerenje kod veine stanovniatva. Vrijeme neizvjesnosti je izazvalo sna~ne i ustrajan posljedice of the war are still present not just in the heads of kod veine stanovniatva, ali i u strukturi budgeta Vlade RH which provides tax relief for war hit area and considerable amount of the budget goes into the war torn area Federal Budget za 2007 (2007). In this climate it is natural for households to have strong preferences towards protecting their savings from adverse effects that might not come from business cycle. All this has left a strong psychological fear in households about the future and might have skewed the expectations towards pessimism. Ekonomski elementi dolaze iz injenice da je Jugoslavija iskusila viae u estalih perioda inflacije. Zapravo je hiperinflacija nepoznata u Hrvatskoj. There was a period of rising inflation and hyper inflation in Croatia from 1990 to 1994: this period of hyperinflation and its subsequent demise was described in Rohatinski at all (1995). The period of very high inflation was from 1992 till 1994 ending with the average inflation of 24% per month. Once new currency Kuna has replaced the Croatian Dinar the inflationary process was gone and price stability ensued. U toj knjizi Rohatinski et. all (1995) provide a great insight into how a period of hyperinflation started in Croatia and what steps were taken in order to curb down the inflation and stabilize the economy. One characteristic of Rohatinski at all. is that the book has several segments where the economics of inflation is replaced with the psychology of inflation. This non monetary part of inflation is hard to quantify and it is solely based on the expectations of economic development and the credibility of monetary and fiscal policy. Psychological element of inflation according to Santini (2007) never left Croatian economy and is still embedded in the economy through cheap imports. Following the rational of Santini (2007) one of the main reasons why Croatia has low inflation and stabile prices is the availability of cheap imports used to meet domestic demand driven by cheap loans. However some economic facts are absolute. The two basic economic indicators in Croatia have been positive for last ten years. The average GDP growth was 3,6% and the average inflation was 3,5% in the period from 1998 to 2006. We can argue are real GDP growth rates "good" or "bad" should the real GDP rates be higher or lower, but for the purpose of our analysis we have to observe the real GDP has positive growth rates in all but one of the last nine years. There was only one negative rate in 1999 when the real GDP growth rates was -0,9%. Although the data does resemble some business cycle properties, one year negative GDP growth rate can hardly be called a recession or an economic demise expected to be tied in with major currency depreciations. Similar analysis can be made with inflation. Od kada je Hrvatska kao mala otvorena ekonomija it is somewhat normal to have more volatile rate of inflation then large economies like USA and Euro Zone, however average inflation rate of 3,5% in the time period form 1998 to 2006 can not be considered disturbing, troublesome or damaging to the economy. Slijedea varijabla u order to determine the soundness of economy is certainly the unemployment rate. For our particular analysis we are going to look at the number of employed. In 1998 on yearly average there was 1,38 million employed in 2006 there were 1,48 million employed as per yearly average. So the employment rose about 110 thousand and the population fell by 60 thousand in the same time period. So the absolute number of employed increased as well and the percentage of employed in the economy. All this data points to an economy that might not be at an optimal path, but it is hardly any evidence of economy that is structurally unsound, experiencing problems or is on the brink of recession/crises. There are some limitations and possible warning signs to Croatian economy like the level of foreign debt, the level of household debt, need for faster and more efficient judicial system. However which economy is not battling with these problems? Also it should be noted that level of foreign debt or government deficit are hardly topics which are discussed in average households or topics of extreme concern to an average household. To nam namee opet pitanje: zaato stanovniatvo zadr~ava atednju u devizama? Jedno od populisti kih objaanjenja je da e te aj devalvirati jer je zemlja inherentno nesatabilna. U periodu od sredine 1998,. godine do prosinca 2007. godine, te aj HRK je bio vrlo ograni enog raspona. Vei dio promatranog perioda je bio 7,28, sa standardnom devijacijom 0,33. Minimum u promatranoj seriji je 6,61 a maksimum 7,73. The lower band is 0,67 Kuna away from the mean or about 9%. The upper bound is 0,45 Kuna away from the mean or 6%, essentially indicating an upward resistance. All this indicates an mean reverting and stabile time series. There were no major instances of depreciation, only one instance of appreciation. The appreciation occurred in 2001 when the new CNB governor came to power. In 2001 there was a sharp move in the exchange rate from 7,6 to 7,4. Since then the exchange rate has been between 7,3 and 7,5 without any clear trend, but a flat line. The exchange rate is not fixed, but it is been kept in a very narrow band. What the data tells us is that Kuna has been extremely stable, almost at the point of fixed exchange rate regime in last seven years. There were many calls for the change of the monetary policy towards more flexible exchange rate regime, gradual or quick depreciation, but those economic arguments were never brought to fruition. Most forceful argument for change of monetary policy was in Santini (2007) On many occasions, the governor of HNB as well as other employees of CNB, have stated the exchange rate stability is the main course of action for the CNB and that there are and will not be any changes in the conduct of the monetary policy or the monetary policy regime Rohatinski (2007). So from the perspective of the average households the argument for holding their savings in foreign currency was never explicitly stated by the governor of CNB or any other public official. If anything the governor constantly repeats his firm stance of the exchange rate stability. Kako je prikazano tom kratkom argumentacijom, Hrvatska ekonomija je bila stabilna u posljednjoj dekadi. There are some memories of violent past both political and economic, but no major socks have occurred in more then a decade, yet Croatian households prefer to save in foreign currency. Predmet ovog rada je osvrt na taj paradoks i pokuaj objanjenja tog fenomena. Ovaj rad je strukturiran kako slijedi, u nastavku je kreiran dinami ki model stanovniatva, drugi dio u slijedu prikazuje model tr~iata kapitala te kreira Bayesian objaanjenje o ekivanja stanovniatva, a etvrti dio prikazuje pregled i implikacije modela i donosi pet zaklju aka. 2. The dynamic model We are going to look at a representative household. We shall assume the household lives infinitely and it is trying to maximize the utility. The utility comes from consumption, denoted as c. The household is a forward looking one so it ties to maximize the expected utility in the future. The future is discounted with the factor  and E denoted expectations.  EMBED Equation.3  We in order to simplify the model we shall assume there is no credit and the household can either save or consume the income. Income comes only from wages, for simplicity and without loss of generality we are going to assume the real wage is constant through time. The household is faced with the budget constraint in each period. It can not spent more then the periods income and existing savings. Here again we are going to eliminate the possibility of credit, since we are investigating the problem of savings and not of consumption or from what source that consumption is derived. The problem of household consumption in Croatia is analyzed in Vidakovic (2005) and Vidakovic (200b) The amount of spending is strictly determined by the household and the households preference for risk. The utility function is  EMBED Equation.3  Where a is a constant and c is consumption. This form of the utility function is the exponential utility. In our case we are going to use it due to its important property of constant relative risk aversion. Using standard measure of risk aversion as specified in Kimball (1993) we have  EMBED Equation.3  The importance of this particular specification is immediately obvious. We are looking for explanation why under any economic activity the households have preferences to keep their savings in foreign currency. So we have to have imposed restriction on the household's behavior in the form of the utility functions. This particular form guarantees that given any economic circumstance the household has the same risk aversion. Before we set up the Bellman equation we need the value function. The value function takes the form:  EMBED Equation.3  As noted in each period the household can choose between savings and consumption. Also the household has past savings noted as A, Y is income in that period and C is again consumption, r is the rate of return on savings. In that case the level of savings in each period is:  EMBED Equation.3  So this period of savings is the accumulated savings from past plus the savings from current period. We are now going to extend the model by adding additional savings options. We are going to split the savings between risky asset and less risky asset. In our case the non risky asset will be savings in foreign currency. The households can obtain the foreign currency through and FX transaction at no cost and in the unlimited amounts. Getting the foreign currency and depositing it in the bank with the rest of the accumulated savings. However if the household has different preferences it can leave the savings in local currency (some or all of it). The savings in local currency we are going to treat as the risky investment. Since the only risk in the model comes form the possibility of the depreciation. The amount of savings invested in risky and non risky asset is determined by the factor  which has the range from 0 to 1.  EMBED Equation.3  In equation 6 z is the rate of return on risk asset and  is the portion of savings invested in risky asset. Now we can set up the bellman equation:  EMBED Equation.3  First order conditions are For c  EMBED Equation.3  For   EMBED Equation.3  In this equation we have used the fact that  EMBED Equation.3  is known at time t. By budget constraint and using envelope theorem we get:  EMBED Equation.3  Eliminating V'(a+1) we get  EMBED Equation.3  Substituting two conditions become:  EMBED Equation.3  This solves the model. The model presented here is a standard dynamic households model that can be found in Sarget and Ljungqvist (2004), Cooper and Adda (2003), Blanchard and Fisher (1989). What we have done in this particular case is allow households to choose between two forms of savings, risky savings in local currency and risk free assets used for savings in form of foreign currency. So in this particular set up the only risk as presented by the expectations operator is the FX risk of domestic currency depreciation. We have set up a dynamic model of a household with saving preferences. In next part we are going to look at the way a household makes decisions when it comes to the choice of savings. 3 Expanding the model As we have seen in the equation 6 the household has two options for savings. The first option is savings in foreign currency which is considered a risk free asset and the second option is to save in local currency which is considered a risky asset. The choice between the two is given determined through the probability and expected return. In this part we are going to look at two particular cases. The first case is going to be the determination of probabilities of investing in each asset and the second one is going to be the impact of the capital markets on the choice of investment. In this part we are also going to determine the difference between risk and uncertainty. As defined by Frank Knight and re-iterated the in the tour de force paper by Sargent and Hansen (2000), risk is the objective probability distribution and uncertainty is the subjective probability distribution around event. The basic example is as follows: a player is faced with a guessing game of guessing what is the color of the ball to be drawn from the bowl; there is a boll with 3 blue balls and 7 red balls. For every correct choice the player receives an award of 1 and for every incorrect choice the player receives the award of 0. The game is played either once or sequentially. In this game the player is faced with a choice, the object of the game is to maximize the profit from playing the game. What are the probabilities of each of the events? The objective probability distribution is 7/10 for red and 3/10 for blue. In any given circumstance any rational player would always bet on the red ball. Considering there is much higher probability the red ball is going to be taken out of the boll. The subjective probability or uncertainty is based on personal bias of the player that is determined with everything, but the number of balls in the bowl. The subjective distribution might be determined by the player's preferences for blue balls. Blue might be her favorite color, or today the player is wearing blue, so play blue. Also the preference might be determined by the prior distribution, the last time the game was played the blue came out so play again the blue. The closest example of choosing subjective over objective distributions in our model of household's savings might be the fact the player (household) has mistrust in the rules of game and the credibility of the set up and the rules of the game. The player might not believe in the set up of the game or has belief the supervisor of the game is not telling the truth. Or the player might play absolute strategy after it was cheated once. In the example with red and blue balls if the player would choose the blue ball based on the subjective distribution over the red ball from based on the objective distribution the outcome of the game would be suboptimal for the player. 3.1 Bayesian probability In order to determine the choice of  for particular period the consumer is faced with the following questions: what is the probability of depreciation of a currency in any period? This is the fundamental question considering the fact the main reason why the savings is kept in foreign currency is probability of depreciation of domestic currency. We are going to set of two likelihood functions; one for each of the events. We are not going to specify exactly what are the priors and posterior at this time. The detail procedure on the Bayesian econometrics can be found in Hamilton (1994) and Koop (2003) The first likelihood function is going to be:  EMBED Equation.3  This is the objective function used in order to determine the probability of depreciation of currency in any given time period. The vector y contains all of the relevant economic information known prior to time period t,  is standard deviation.  EMBED Equation.3  The second likelihood function represents the subjective probability of default and the  presents the subjective vector of information. Once the p and q for each time period are calculated we are going to use the Bayesian model averaging in order to get the true model used in each time period. Following Sims (2003) we get the following formula for :  EMBED Equation.3  The parameter  has values 0 d" 1 and it represents the weights assigned to each of the models. We are going to assume the value of  are fixed for all time. So once the weights between the subjective and objective model are chosen they remain unchanged for the rest of the time. This type of behavior can be seen in the data. The time distribution of savings deposits between the domestic and foreign currency. What household effectively does is to average the subjective and objective model. In the case the objective model has higher weight in the eyes of the household the choice of investing is always going to be on the mixture of the risk free asset and the risky asset. In case the consumer has bias towards subjective model the choice of savings is going to be the risk less asset. This preference, the value of  is determined by the risk aversion defined in part 2. In case the value of =0 we have the absolute weight on the subjective model and the household will only save in foreign currency. This in effect explains the paradox of Kuna savings and why Croatians save in foreign currency. Croatians simply choose to ignore the objective model and have preferences towards the subjective model or some mixture of subjective and objective model. The choice of subjective model is what leads towards the suboptimal choice of savings, constantly preferring lower returns on foreign currency savings then investing in the mixture of risky an risk less asset in order to maximize their return. 4 The capital markets In this part we are going to investigate the choice of the structure of the portfolio if we substitute the risky asset with investment in the stock market. In the model so far we have investigated the choice of the investment in a bank deposit in foreign and local currency. Now we are going to investigate the choice between investment in bank deposit in foreign currency and investment in stock market which is in local currency. The likelihood functions are going to remain the same, except we are going to modify the parameter  in the model averaging. The theoretical rational for this assumption comes was originally proposed in Lucas (1972) and it was empirically investigated in Lucas (1973). In Lucas (1972) author creates a model where the customers are rational, however due to miss information there is a possibility of economic participants reacting to nominal disturbances. In a series of papers Sims (2003), Makiew and Reise (2002) and Reise (2006, 2006a) investigate behavior of economic participants that are information based. In these models the economic participants collect information; however they choose to act up the information collected only after certain threshold of attention. In this part of the paper we are going to analyze the behavior of households and how do they react to high returns in stock market. We are going to assume there is some threshold of attention for the parameter  where the subjective probability model is rejected in favor of the objective model. The only scenario where this would happen is when the return on the risky asset is so substantial that it is larger then the risk aversion factor a. So if we have a set up of  EMBED Equation.3  In that case the expected return from investment in the risky asset is greater then the risk aversion factor and the household is going to reject the subjective distribution in favor of objective and redistribute the investment portfolio between the risky and risk less investment. In essence we have not allowed the households to change the parameter  over time as stock market investments are available. The implications of this setup have far reaching consequences. It is obvious that unless the passive bank rates are very high in a stabile market conditions the households are not going to change their savings from foreign currency into local currency. Also the rise in stock market prices is going to draw savings from banks into investments in stock market. 4.1 The cost of risk aversion The graph below provides us with the average Euro interest rates and HRK interest rates for households and their differential.  Source: CNB and author's calculation As it can be seen from the graph above the HRK rates have been persistently higher then the Euro rates in long term household deposits. On the graph we can also see the differential between the Euro and Hrk rates, which just quantifies what, can be seen on the graph. In order to determine the actual cost on 1 HRK deposit. We have used the following formula:  EMBED Equation.3  Where DEP is 1 HRK of deposit i are the HRK and Euro interest rates. Once the calculation of the equation 15 is performed the value we get is 1,42. What this effective means is that in time period from 1995 to 2006 Croatian household has forgone 42% in interest on their deposit because of their preference to save in foreign currency over local currency. The above calculation is astounding to say at least. This value is basically the cost of using subjective over objective probability distribution when making household decisions. However as it can be seen the level of 2,7% a year of foregone income is obviously not enough to move households from savings in FC to savings in LC. Now we can see the impact of the capital markets on savings in LC. We are going to use CROBEX index as the main indicator of returns on the stock market. The data for CROBEX shows that the return on COBEX from its start on 02/09/1997 with the value of 1118 to 31/12/2003 with the value of 1185 the return in stock market for the period has been fewer than 6%. After 2003 we see rise in the index. On 30/12/2005 the index closes at 1998 after breaking 2000 in value in 2005. From that period on we see an explosion, the index ends 2006 at 3209 and 2007 at 5239. The return on index in 2006 was 60% in 2007 was 63% and the cumulative return was 162% in two years. How did the households reach to these returns? The best way is to look at investment funds industry which has boomed. The net assets of open investment funds grew by 81,5% in 2006 and 87,3% in 2007, far exciding the stock market returns. In part 3.1 of this paper we have looked into the possibility there is a threshold of attention where the households are willing to invest into what they perceive as risky asset. Given the nature of the model it is hard to quantitatively determine what the level of threshold is and for now we are going to leave that to future research. However the empirical fact does remain the households are willing to switch their preference of saving from foreign currency to local currency if the level of incentive in terms of returns is substantial enough. This was seen in high levels of inflow of money into investment funds. 5. DKM . / 2 4 6 X Z PRzfzWjh}ohs%gUmH sH 'jh}ohs%g0JUmH sH h}ohs%g0JmH sH !jh}ohs%g0JUmH sH h}ohEmH sH h}ohs%gmH sH h`mH sH h}oh`mH sH h}oh*&mH sH h4mH sH h}oh8=mH sH h}ohE5CJ aJ mH sH h}ohs%g5CJ aJ mH sH CDLM   4 \ j l xgds%g $xa$gds%g $xa$gds%g  0>@BR^6:<Bjnt(*>BƻƇ|u|nnnn||nn||| hnhQ& hnhP-h2 hnh2 hnh]h}ohy5CJaJmH sH h]5CJaJmH sH h}ohs%g5CJaJmH sH h}oh*&mH sH h}ohs%gmH sH h}ohs%g0JmH sH jh}ohs%gUmH sH #jh}ohs%gUmH sH *   PB#%#''*+-/124J68:8=S>?A $dha$gd!sxgds%g>HLNRVX`|&*0DPV*,.RT^pzʶʯʨèäèè h ahP-ht h ah. h ahyjh ahP-0JU h ah}P h ahn hnhyh\+G hnhP- hnhn hnhQ& hnh]h2hn> :@bfZ^`bnpr|0<h~  "(*2^` *,6> h(h( h(hP- h(h}P h(hQ@ h ah}P h ah(h( h ah.ht h ahP-L>@Vlprzdfvxz| <BHPTVln~ûʷûåûûh}oh}PmH sH  hRh}PhRhRmH sH  hRhR hRh(h(mH sH  h(h?5 h(h4 h(h}P h(h(h( h(hP-A  (*.2:<@PRTXZ\lnrt|~$&,.68<@BFJZbf|h%h}ohP-mH sH  hRhP-hR hRh}Ph}oh}PmH sH hRmH sH  hRhRJ  68JLVXZ\^vx~   & ( , . 4 6 < > D F źź񺳺񫳧h}oh}PmH sH h%h%mHsH h%h}Ph%h%mHsHh%h}PmHsH hRh}P hRhP-hRmH sH h}ohP-mH sH  h%h%h%mH sH =F P R ` b h j r t x z !! ! !!!(!*!0!2!>!@!F!H!N!P!V!X!`!b!h!j!!!!!!!!!!!!!!!!!!!!!!h%mH sH  h%h%h%h%mH sH  h%h}Ph}oh}PmH sH S!!!!"""""z""#5#>#B#D#H#K#S#T#$2$4$6$F$H$L$|$$$$$$$$$$%%%%%&7&8&?&B&a&e&&"'#'׺גh}ohymH sH h}oh> mH sH hCjmH sH  hRhy hRhR hRh}P hRhQ@h}ohU DmH sH h}ohDmH sH h}oh}PmH sH  h%h}Ph%mH sH h4mH sH  h%h42#'/'?'@'B'I'Q'w'''(((((-(9(A(H(((>)?)@)o)w)x)}))))))***++,,:,A,J,],$--------/P///000 00ڹ幮h}ohmH sH hRmH sH h}oh"HmH sH h}ohrmH sH h}ohymH sH h}ohtmH sH h}oh}PmH sH hCjmH sH h}oh> mH sH hRmH sH <0000 11q11111d2222222222[3|3333455 5|5555556(6D6J6L6N6\6^6b6d6r6|666666ɾɯ hX}hxA hX}hRhRmH sH  hRh hRhxA hRh.4 hRhRhmH sH h}ohrmH sH h}ohxAmH sH jh0JUmH sH h}ohF9mH sH h}ohmH sH hmH sH 46666666666666666677777`7n7t7777777777777888?8888@9D9P9a9!:2:F:H:[:]:::::::;;<<<"=4=w=ɶɶɶh}oh"mH sH h}oh.4mH sH hmH sH h}ohF9mH sH h}ohxAmH sH  hX}hF9hX} hX}hX} hX}hxAhX}mH sH Cw===Q>S>h>i>p>y>>>>>>>>>>?? ??? ?$?6?t?????????????????????????????@@@ @걦h hhJ hh%h}oh}PmH sH h}ohrmH sH hmH sH hmH sH hmHsHhh.4mHsHhhmHsHh}oh.4mH sH h}ohz6mH sH : @@@@*@B@F@X@l@@@@@@@@@@@@@@A A$A2A4ATQTRTSTTTsT}TTTɸ{ꭞiXMh}ohtRmH sH !jh}ohweEHUmH sH #jYAK h}ohweUVmH sH !jh}ohweEHUmH sH #jAK h}ohweUVmH sH jh}ohweUmH sH h}ohwemH sH !j h}ohweEHUmH sH #jAK h}ohweUVmH sH jh}oh9L2UmH sH h}oh9L2mH sH h}oh!smH sH TTTTTTTTTTTTTTUU,U-U.U/U0U1UHUU{iX{MBh}oh=%mH sH h}ohV/mH sH !j(h}oh+IEHUmH sH #jBK h}oh+IUVmH sH h}ohtRmH sH !jL$h}ohA EHUmH sH #jABK h}ohA UVmH sH jh}ohA UmH sH h}ohA mH sH jh}ohweUmH sH !jj h}ohA EHUmH sH #jNAK h}ohA UVmH sH h}ohwemH sH U0U1U@WWWWXX`Z[ ]]`0c1cJcKc]f`gggggHiJizi $dha$gd!s$ & Fdha$gdY+UUUUUUUUHVIV@WWWWWXX2X3XsX`Z[[[[[[[&\Z\\\]{ppeeTe!jh}oh2 G0JUmH sH h}oh2 GmH sH h}ohui mH sH h_b}mH sH h}ohOm%mH sH h}oh=WmH sH hY+mH sH h}oh*&5CJaJmH sH h}oh%w5CJaJmH sH h}oh%wmH sH h}ohCqmH sH h}ohV/mH sH hSmH sH h}oh=%mH sH hShSmH sH  ]^^^K_________e`j``````aa!a5a>aBa~aaaaaaa@bDbgbhbkbbbbbbbcc+c-c.c0c1cJcfߥߥߥh}oh=W5mH sH h}oh=WmH sH h}oh!mH sH !jh}oh0JUmH sH h}ohmH sH h,EmH sH h}ohMmH sH h}ohui mH sH h}oh2 GmH sH h}ohOm%mH sH 4fIgJgNgOgTg_ggggggggggggiDiJiLiritivixiiiĵ|nfWF!j0h#dh?uEHUmH sH j`ÄK h?uCJUVaJh'mH sH jhyd0JUmH sH h}oh%wmH sH !j,h}ohyEHUmH sH #j}K h}ohyUVmH sH h}ohymH sH jh}ohyUmH sH h}ohV/mH sH h}ohCqmH sH h}oh=WmH sH hydmH sH hVmH sH h}oh_3mH sH zi|ijBlDltlvl+rtt5t6tvz||||~$ & Fdha$gdY+ $dha$gd!si*j,j.j0jk>l@lDlFlllnlplrltlvl2mnnnn,o0onowopppq(qPq*rBrbrrrrrr%s'sڱڊڊttltltltttlh)bLmH sH h}ohmH sH h}oh#ImH sH h}oh_3mH sH h}ohJCmH sH !jY4h}oh@EHUmH sH #jK h}oh@UVmH sH jh}oh@UmH sH hydmH sH h}oh@mH sH hlemH sH h#dmH sH h}ohymH sH ('sFsHs`s|sttt t2t5tqt)uuuvvtxxxyy~yy0z>zHzϿ~s~kc[S[c[K[C[hmH sH hymH sH h2mH sH h(O`mH sH h`mH sH h!mH sH h}oh#ImH sH h}ohJCmH sH h}oh{ mH sH h}oh!mH sH hyh@5CJaJmH sH hyhOm%5CJaJmH sH hyh5CJaJmH sH hyhy5CJaJmH sH h}ohOm%mH sH h}ohmH sH h}ohNmH sH HzXz`zzz||||||o}}~~hڼڠڠڕ}qf[fLf[f[j9h}oh6UmH sH h}ohBomH sH h}oh mH sH hyhK5mH sH hyh 5mH sH hyhy5mH sH h}oh#ImH sH h}ohKmH sH !j{7h}oh#dEHUmH sH jAK h#dCJUVaJjh}ohJCUmH sH h}ohJCmH sH h}oh`mH sH h(O`mH sH hmH sH 01KhՋ֋l:R<>Z\^ $xa$gdUb $xa$gd}o$ & Fdha$gdY+ $dha$gd!s,-./1in_;EΈghrsԋ֋؋ɸۭꭚ|tiaiYihmH sH hEmH sH h}oh% mH sH hKymH sH h}ohKymH sH h}oh.O8mH sH h\ mH sH h}ohiIamH sH h6mH sH h}oh mH sH !jtHh}ohBoEHUmH sH #jCFK h}ohBoUVmH sH jh}ohBoUmH sH h}ohBomH sH h}ohRmH sH ؋ً&(6DZ^`n|~,:bfpvzž hlhl hlhe3 hlh%  hlhhPh$ hl hlhK hlhlh}oh mH sH hl5CJaJmH sH Uh}ohK5CJaJmH sH h}oh 5CJaJmH sH 8Zaklju ak Ovaj rad je pokuaaj objaanjenja o ekivanja stanovniatva da se suo i s razli itim ulaganjima i glavnim izvorima rizika koji proizlazi iz te ajne deprecijacije kune. Model prikazuje okru~enje gdje stanovniatvo kreira subjektivnu i objektivnu distribuciju vjerojatnosti deprecijacije domae valute. Objektivna distribucija se temelji na aktualnom ekonomskom podacima i uvjetima; a subjektivna je temeljena na strahu iz proalosti. Nakon koriatenja Bayesian modela stanovniatvo u prosjeku, ima mogunost izbora izmeu investiranja u rizi nu imovinu i nerizi nu imovinu. However the choice of the distribution of savings in each asset is going to be determined by the probability weights assigned to objective and subjective model. Slu aj subjektivnog modela daje veu te~inu nego je potrebno za stanovniatvo za prihva anje viaeg povrata na atednju, even negative returns if the subjective model predicts depreciation. Model then creates an arbitrary threshold of attention for the household where household is willing to take a chance by investing in a risky asset. Empirijski to se pokazuje kao prag pozornosti, kada analizirana razlika izmeu Euro i HRK deposits was not enough for average households to change their savings preferences. Once the stock market is introduced in the model we have seen households do move into risky investments. However the returns on the stock market have been over 60% per year in 2006 and 2007. To prikazuje povrat koji je potreban za seljenje iz devizne u atednju u domaoj valuti, koji je toliko visok da mo~e biti substantialan. Opi zaklju ak ovog rada je da postoji okru~enje u kojem je stanovniatvo sklonije koristiti subjektivnu distribuciju nego objektivnu, pa e ono prihvatiti i ni~i povrat na svoju atednju; ato je ovim radom iskazano kao suboptimalno o ekivanje. Bibliography Blanchard, Olivier; Ficher Stanley (1989): Lectures on Macroeconomics MIT Press 650 str Cooper, Rusel; Adda, Jerome (2003): Dynamic Programming: Theory and Macroeconomic Applications MIT Press 296 str. Christiano, Lawrence: (2002): Solving Dynamic Equilibrium Models by a Method of Undetermined coefficients, Computational Economics, Volume 20, Issue 1-2. Gregurek, Miroslav; Vidakovic, Neven (2007) The transition game  who played it and with what resoults Working paper. Hamlilton, James (1994) Time series analysis Princeton University Press 799 HNB (2007): Bilten HNB-a broj 132, HNB publikacije Prosinac www.hnb.hr Kimball, M. (1990): Precautionary Saving in the Small and in the Large," Econometrica (January), 53-73. Kimball, M. S. (1993)  Standard Risk Aversion. Econometrica, Vol. 63(3), pp. 589 611. Koop, Gary (2003) Bayesian Econometrics John Wiley & Sons 359 Lucas Robert Jr (1972) HYPERLINK "http://ideas.repec.org/a/eee/jetheo/v4y1972i2p103-124.html" Expectations and the neutrality of money.  HYPERLINK "http://ideas.repec.org/s/eee/jetheo.html" Journal of Economic Theory, Elsavier,, Elsevier, vol. 4 2 , pages 103-124, Lucas Robert E Jr (1973)  HYPERLINK "http://ideas.repec.org/a/aea/aecrev/v63y1973i3p326-34.html" Some International Evidence on Output-Inflation Tradeoffs.  HYPERLINK "http://ideas.repec.org/s/aea/aecrev.html" American Economic Review, American Economic Association, American Economic Association, vol. 63 3 , pages 326-34 Ministry of finance Budget of Republic of Croatia for 2007 http://www.mfin.hr/str/10/ Reis Ricardo(2006)  HYPERLINK "http://ideas.repec.org/a/eee/moneco/v53y2006i8p1761-1800.html" Inattentive consumers.  HYPERLINK "http://ideas.repec.org/s/eee/moneco.html" Journal of Monetary Economics, Elsavier,, vol. 53 8 , November, pages 1761-1800, Reis Ricardo (2006a)  HYPERLINK "http://ideas.repec.org/a/bla/restud/v73y2006i3p793-821.html" Inattentive Producers.  HYPERLINK "http://ideas.repec.org/s/bla/restud.html" Review of Economic Studies, Blackwell Publishing, vol. 73 3 , pages 793-821, Reise, Ricardo Mankiew, Gregory N. (2002) Sticky Information Versus Sticky Prices: A Proposal to Replace the New Keynesian Phillips Curve. Quarterly Journal of Economics, November vol. 117 4 , pp. 1295-1328,. Rohatinski, }eljko (2007): Euro Teze za izlaganje na forumu Udruge "Napredna Hrvatska HR+" o otvorenim pitanjima uvoenja eura u Republici Hrvatskoj, Zagreb, 19. lipnja 2007. www.hnb.hr Rohatinski }eljko, Anuai Zoran, `onje Velimir, (1995) Put u nisku inflaciju Hrvatska 1993-1994. Vlada Republike Hrvatske , Zagreb 171 pages Santini, Guste (1994) Ekonomska politika za 1994. godinu Rifin Zagreb Santini, Guste (2005): Gubici izravnih poreza temeljem deficita robne razmjene s inozemstvom u Hrvatskoj za razdoblje 1994.-2004. godine Ekonomija/Economics godina 12 volumen 1 Santini, Guste (2006): U inci te ajne politike na gospodarstvo Hrvatske u razdoblju 1994. - 2005. godine Ekonomija/Economics godina 13 issue 1 Santini, Guste (2007): Iluzija i stvarnost hrvatskoga gospodarstva Rifin Zagreb 278 str. Sargent, Thomas; Hansen, Lars Peter; (1981): Linear Rational Expectations Models of Dynamically Interrelated Variables, in: Rational Expectations and Econometric Practice, University of Minnesota Press Sargent, Thomas; Hansen, Lars Peter (1991): Two Difficulties in Interpreting Vector Autoregressions, in Rational Expectations Econometrics, ed. by Lars Peter Hansen and Thomas J. Sargent, Westview Press Sargent, Thomas; Hansen, Lars Peter; (1991): Identification of Continuous Time Rational Expectations Models From Discrete Time Data, in Rational Expectations Econometrics, ed. by Lars Peter Hansen and Thomas J. Sargent,Westview Press Sargent, Thomas; Hansen, Lars Peter; (1991): Exact Linear Rational Expectations Models: Specification and Estimation, in Rational Expectations Econometrics, ed. by Lars Peter Hansen and Thomas J. Sargent, Westview Press Sargent, Thomas; Hansen, Lars (2000) "Wanting Robustness in Macroeconomics" working paper Sargent, Thomas J. and Ljungqvist, Lars (2004) Recursive Macroeconomic Theory, Second Edition, MIT Press, Sims Christopher (2001) Solving Linear Rational Expectations Models. Journal of Computational Economics, 20 1-2 , p.1-20. Sims Christopher (2003) Implications of Rational Inattention. Journal of Monetary Economics, Elsavier,, 50 3 , April. 665-690 Vidakovic, Neven (2005): Theory of Rational Expectations and Microeconomics of Debt Ekonomija godina 12, broj 4 705-729 Vidakovic, Neven (2005b): Model of Economy without a banking system Ekonomija/Economics godina 12, broj 3 551-569  `tednja i depoziti, uklju ivo avista atednju. Podaci iz Bilten HNB-a  The data comes from www.hnb.hr and the CNB Bulletin  The wrong answer has the reward of 0 in order to avoid the gambler's ruin problem.  Something like "sure they say there are 7 red balls, but the government never tells the truth"  The likelihood function does not give us the probability just the expected exchange rate, which is then compared with the period's exchange rate in order to determine will there be a depreciation.     hl&|~BZ\4:<n PVǿǿǿǿǷǷǬǤhcb@mH sH h\mH sH h}ohWmH sH h$ mH sH h}oh% mH sH hlmH sH hhbamH sH h}oh0mH sH h}ohe3mH sH  h9Ohe3 h9Ohl hlh%  hlhT~ hlhl38<Vbdt 0HJLPR "2<DPVXZ\^hjr48ƿƿƸƿƿͱ͸ͱͱƸƸƸ hh9O hhhba hh$ hh}R hh!s hh hhcb@ hh%~hh}ohWmH sH h@mH sH C8:<>XZ\^FHX2,,:>FHtBǿ翎xmeZh)h)mH sH h2 mH sH hVh@mH sH hVhVmH sH hVh2 mH sH hVmH sH hUbh}omH sH h@mH sH hUbhUbmH sH h}ohUbmH sH hUbmH sH h}omH sH hUbh}o5mH sH hUbhUb5mH sH h}oh}omH sH  hhJC hhW ,B>dZxBf $xa$gdCj $xa$gd/ $xa$gd) $xa$gdUb $xa$gd}o02:<<>\`hjlNPTV(*rʿwh xh xmH sH h}omH sH h xmH sH jhUbh}oUmH sH hUbh@mH sH h@mH sH hUbh}omH sH h/hVhmH sH hVhmH sH hY+mH sH h/mH sH h/hY+mH sH h)h)mH sH h)mH sH -fhXZln BDZ.2:<퟉h4hUbmHsHhCjh@mH sH hCjhCjmH sH hUbmH sH hUbhUbmH sH hUbh}omH sH h@h@mH sH jh}oh}oUmH sH h}oh@mH sH h@mH sH h}oh}omH sH 14flnvz  ^LN̶̶̘̣̘̣yh!jhydh#a0JUmH sH  h ah#ajh ah#a0JUhUbh}omH sH h}oh@mH sH h}oh}omH sH hSmH sH hY+hSmH sH hY+h@mH sH h@mH sH hY+hY+mH sH hY+hUbmH sH hUbhUbmH sH h}ohUbmH sH $fLf( $xa$gd}o $xa$gdUbNfh(*h}oh}omH sH h#ajh#aUhIQf!jhydh#a0JUmH sH hydh#amH sH 21h:p!s. A!n"n#n$n% DyK neven.vidakovic@pbz.hryK <mailto:neven.vidakovic@pbz.hrDyK neven.vidakovic@pbz.hryK <mailto:neven.vidakovic@pbz.hrDd b  c $A? ?3"`?2aA؍ϘSn_>5=`!5A؍ϘSn_>5N  &@2xڥT=H#A~vc&b. bEE9FlXKT" #N'Wr 66vjqX+TP1뼙UCV#2[|{>AXA'VC۶e>oA7/`BYDD%P 6%CXଈʆ GkzO$ve;厤?T(\~̼X̎5gV]*VNƷc@'JE]N}SAa:=,]xi@i F}(n¿5x)|Ýw'n§nq$I{o&u;ޡ}ײΥ/qX]lbY_`~­`:#D*-~O?}Ahw}ZgpT͜\7@#,~Ү[~}dǟbt}jp S./]x'OKNws' ~<;cu1J'G v¬@ WѐmjDd t|b  c $A? ?3"`?2ЧQ!ɏ8 t촭K`!ЧQ!ɏ8 t촭 ` 0Vxcdd``ed``baV d,FYzP1n:&v! KA?H1ā깡jx|K2B* RvfRv,L ! ~ Ay +?ObהҒXt96!|&]&I%g>#g^ 0~3d8_/MF(HP 27)?!\uG!_+] -͵aa\ս W_. pH悦.pJ[M `PDcdbR ,.IeHԡ"|b@3X?Ǒ1Dd b  c $A? ?3"`?2XpzPNj͸*K`!XpzPNj͸*.@e xڥR;KAٽ΀BLᣈe@b/ZDNpVJb */'k! 羲1||7)B4c" *IE[Iu6yij@fl| d2Y7FYaCIvNd_r/üAN>%\(4X}ϭ;n٬y3gqleNm瑕P&Rw45t'"zi]O4x@չǂ£}>zu_ؿ-MX 8~~T_@88xN =/(l2 { !P3-u[oQy4PDd b  c $A? ?3"`?2lKC/b`%vv K`!nlKC/b`%v8\ @2<xڥ=hA߼3] l,,$ha>Len N1AT i$uX)!D b̛/[=v~˛c0QA+Г-QS=AYvT}kq)cޙCi8CFpYDu 1FŤ[=r-Ll/Ӥ(~^8o[k6<~Qz!ZgU);Etxw[7e mCz ^IsM WFO4?y`Sc(h40=.T90i9$C~xK< 7'ī:>o:>y"|$,x/į#:VѱHF\d Lw}o k^)w}w֙KYR(j^7ܓ!k$8_{_~Ǯs}JWrWp7k6t{$>d-8ǓR =osk4w<~>Lj%4@U 4HZOBatA8 O XذCp Dd ( b  c $A? ?3"`?2BYFb{tq<ZTh K`!YFb{tq<ZT@1xڥKAg.K41 8⣲"FB!`T4Q1vZ[)V?`HD;{5w&(egY@ڠ"?{}QBP0xN"/@rO("z00B/Xt *=VN' }! ?p{aQʞаekA%~kfj=㩞:aƲ۩,nm*7_D-.L71J.CɌOMU%A^dIFnx丟~"ߺKr!Y_\{/_ |] ?U-5r3}Rr~wlu[xi[oH$h~k l_׈9Ks|s}9T#/ VxR>c0Yn0[wG?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~(Root Entry_ FH2Data :KWordDocument^4ObjectPoolawHH_1265098677@FwHHOle CompObjfObjInfo !"#&)*+,/234569<=>?@CFGJMNQTUVWXY\_`abcdefgjmnopqruxyz{| FMicrosoft Equation 3.0 DS Equation Equation.39qb0|Z E1+() "t u(c t ) 0" +" 0[]Equation Native _1266643798 FH2HOle CompObj f FMicrosoft Equation 3.0 DS Equation Equation.39qEDX(L[ u(c t )=e "ac FMicrosoft Equation 3.0 DS Equation Equation.39qObjInfo Equation Native  `_1266643939F2H2HOle  CompObjfObjInfoEquation Native _12650986821EF2H2HEfL[ "u''(c t )u'(c t )=a FMicrosoft Equation 3.0 DS Equation Equation.39qOle CompObjfObjInfoEquation Native bt\ V t (A t )=maxE1+() "t u(c t ) 0" +" t[] FMicrosoft Equation 3.0 DS Equation Equation.39q_1265099075F2H2HOle CompObjfObjInfoEquation Native  _1266651382OTF2H2HOle $CompObj %fbзL[ A t =A t"1 +(Y T "C t )1+r t ()Y t "I t ,A t e"0 FMicrosoft Equation 3.0 DS Equation Equation.39qObjInfo!'Equation Native ()_1262615505$F2H2HOle -E  *L[ A t =A t"1 +(Y T "C t )1+r t () t +1+z t ()1" t ()[] FMicrosoft Equation 3.0 DS Equation Equation.39qCompObj#%.fObjInfo&0Equation Native 1M_1262615572",)F2H2H]10B< V t (A t )='max c t  t  u(c t )+1+() "1 EV t+1 (A t+1 )t[]{}Ole 7CompObj(*8fObjInfo+:Equation Native ;z FMicrosoft Equation 3.0 DS Equation Equation.39q]^D u'(c t )=E1+() "1 1+r t () t +1+z t ()1" t ()()V' t+1 (A t+1 )t[] FMicrosoft Equation 3.0 DS Equation Equation.39q]£1D EV' t+1 (A t+1 )(r t "z t )t[]=0_1262615794.F2H2HOle ACompObj-/BfObjInfo0DEquation Native E_1262615897';3F2H2HOle HCompObj24If FMicrosoft Equation 3.0 DS Equation Equation.39q]u8< (1+) "1 (A T +Y T "C t )ObjInfo5KEquation Native L_12626163988F2H2HOle OCompObj79PfObjInfo:REquation Native S_12626166416=F2H2H FMicrosoft Equation 3.0 DS Equation Equation.39q]„ V'(A t )=E1+() "1 1+r t () t +1+z t ()1" t ()()V' t+1 (A t+1 )t[]=u'(c t ) FMicrosoft Equation 3.0 DS Equation Equation.39q]<pA4 u'(c t )=E1+() "Ole ZCompObj<>[fObjInfo?]Equation Native ^X1 1+r t () t +1+z t ()1" t ()()u'(c t+1 )t[]E1+r t ()u'(c t+1 )t[]=Eu'(c t+1 )1+z t ()t[]_1262616761BF2H2HOle hCompObjACifObjInfoDk FMicrosoft Equation 3.0 DS Equation Equation.39q]Ÿ u'(c t )=1+() "1 1+r t ()Eu'(c t+1 )t[]u'(cEquation Native l_1266646523 GF2H2HOle sCompObjFHtf t )=1+() "1 E1+z t ()u'(c t+1 )t[] FMicrosoft Equation 3.0 DS Equation Equation.39qE/L[ f p tObjInfoIvEquation Native wK_1266991968LF2HHOle } p t"1 ....p 0  .  =1 2 2 exp"(y t "c"y t"1 2 2 [] FMicrosoft Equation 3.0 DS Equation Equation.39qCompObjKM~fObjInfoNEquation Native K_1266647951QFHHu/PGL[ f q t q t"1 ....q 0  .  =1 2 2 exp"(d t "b" t"1 2 2 []Ole CompObjPRfObjInfoSEquation Native 2 FMicrosoft Equation 3.0 DS Equation Equation.39qEPGL[ =f p t p t"1 ....p 0  .  +(1")f q t q t"1 ....q 0  . _1266910273YJVFHHOle CompObjUWfObjInfoX FMicrosoft Equation 3.0 DS Equation Equation.39qu#.L[ z t >a FMicrosoft Equation 3.0 DS Equation Equation.39qEquation Native ?_1266763331[FHHOle CompObjZ\fObjInfo]Equation Native 1Table6vSummaryInformation(`$u¦(t[ DEP=(1+i hrk "i euro ) i=1995i=2007 "Oh+'0(4LXl    _J'Sp>hMsf9#|<IDd Xhb  c $A? ?3"`?2| U>*Fo`K`!g| U>*Fz@X+|5xcdd`` @c112BYL%bpu=e]U5c@ϊOUIĿѮ7?A@%K4J5M4xx$7׆nwt[w%FMuCu7-K݅*86,tG vN/ig F~MJݢNSs9-:k/0x2im 3D?!2y@}B08\8;%7-~aJ4(&w$eNe^L0~gx<@1Iһ|wJѸݮ*}sC;cѝ P^ͅ7O]qݧA~;.%ffe~4wMktg3~w~}k{~6~)c2e5܁!v\Ʃa~a&s3N-Oh/J$ig]37ukq*q[Cݡst^9.*GQ~+h4ooo#cvP¹EΪ~γi@R<TQLҹA\{=~TGO}㾈Y9FjY|+z]ָtpB"?oy۞FL ErXW/hqةm<F90N|!! Ȉ8\gĜda.8)}/q{k{Sɑ7q%i^iT6h_1Ԥ=kI/iDd b   c $A ? ?3"`? 2,/&GzB^& K`!/&GzB^& (>xڥMLAL[nZdV 58MEf=Jz7z1@fb bSL5&PCgV7og!lQ Ik#JHuDhRQGIǹ)Qg"P`2^`<.tYQe* Ss慎&QR'tlg<_VâyDJvڭl- |I(_1s&lS%\wm/1]Ǵ߻-M{?CZ9b֭ifݚ_2v&^5tc8 <]i7ԊmkAsf6n6uvPͮš.XTrZJf ~ $޵3wS]xe2*D>̹/ cg"?1Q"W/~'+Gi09Xl}$NTH6GE>#lXȖW Z%Nt \(Zg9u/qۑoq~oC+Iݖ؈JPr惀|/Pb{Cg\n:/4x_su]ɭ̌n[یr3z{|ze3zp n3_2%LJ {؃Xłuz ,Ҁ2z <؈͛ h癋y^( &U%gv.FB-ю.NOOq>KxD7x"p{ͧ¡ ͧ¯4 4 3UX]HQdA|ϡ98PDbqhoY|.,d-庹̽ a'i{ #OUI~gd;Mׅz/}kq9ЖD,QR 0%2ĢDDd Hb   c $A ? ?3"`? 2Gߤ)JR#C)K`!ߤ)JRz @+ "xڥV=LA~on~ Ĝ&@eCr0 !1LDr5ՆRքZЍ!4XЙX B(sfvwY9d/;Λfޛ"DBav^kB-赐4 .]\x5dS-18ë;b06]Ky r$ExxE ­Vi# mfh~FYƞ#+̆A|5ߜ?7z8cϧˋw;bx:\ n&9KTR"9RIgng{ -NN1iV:xnDҜI.~S6[o~ڍVk׭lhɫdR#׼D,yYwbf뙫*]Nq3_q|WbޓxGo>~.jz~*Q+zz:eN_ Sn񇨎`>lM='9~w{s+ͣ91;8;/X~0p|INjef=|ϓwdZBg@XgoVP7ߣOz::U:ע_"O2/jz~ϗ'b+-K[W%DyVou$ĸ] u8bMm%!F!1k EDd b  c $A? ?3"`? 2 R| o^i@-K`! R| o^i``(V'PxڥUMkQI3&~AJ0F\;PA[DPȄf\u! q]PpN. ?Zhs md=w;sd#8[*^ǣJ\%pI+eR4:=#cd0)BO2G5Fr0]/z5`b4ۡj:!>ۈE"i#5C\$/ȓ>$kXr=ܤ֍ugs9ܷY԰pe*w7|涨`ϜTKSf%vc)[YM%li]"v 6!gְ_3S< ،LA ? Dd ,b  c $A? ?3"`?24GT58Q0K`!4GT58Qp((+xڥkAl$d[5)"QzIKA1 DbMFXpm=YRţ )RW,x&P%zfCZ*a>oE` `#D/FH,ljlpǷE3꺅bKLvi6}&:v6xΐ;:XTnWc5JN\]b]c<_Ek7v_*f/{elN@2?UGjMb+> '>%G{~Bn:n9գHG ¿;:AƇ]jٰ,L_Л3L8'"Dd  b  c $A? ?3"`?2lݛ@e9qҩLH4K`!@ݛ@e9qҩL`` hxcdd``>$d@9`,&FF(`T)AXRcgbR 0P7$# ( (,@1[ɔQfg@babAn͔{# &br<'մ{q1]iի|8oX SÁ^U}D=} d H$0ul0] ]L0i`O 75H9>}/CYQ,>zZa(|\iDGdEa㉲j`G<^0CX W&00r]Qh #%ݡؓ=FQ׆C[82k`27n &T_-ɸ81@"+KRsePdk)Y= 7L 1Dd hb  c $A? ?3"`?2{6CUfJruW7K`!O6CUfJru @d|xcdd`` @c112BYL%bpupk0}{8о?0>+ntb@g9t8\q7}L_'O@gz8tGprz\N;`};lrX}2̲9=+vhs,Gnd-coilخamQiD=9<PG1,|ͿF:K1xuϫGqnXx1n{ [~]?irM(4m?t<^"sy}~pB}LumFZ\6[g[i/:z6.>GYáVV$tٞҴM^ۯNį#nHdS o΃k7,N:@F MwL׏dȕV}ڜfm{JNn{!}H2εtu[sIϓ&H{F'1yv2 1nG;L/bc^\Eozl4VȬ^/B\]g1}XE|-c@e^hZ0La;TjQ5E5qy$5j?!HWa8x^UߪBu^S_HjԾO)Ua<5V:U3JU:7Iڟ u1W9ZdUN;Ԩ} ocb^OuakAk7E88F>61ݳc;0XԨ%4.b|71ָ vjA,Hjl q kjhlDMWqGS~v:Ѽg@Cky>Z{Q0~iADkQբӃOW/|Zj#AZGd1ZNfn,ZT`քEi!ߚƼgEQ^ B>?qKD -+1^e0xOD5 -xak(jAZ݉9Z:e5|炜M"eׂc[`{Ӟ¶g xsȋonٌ(E}L컟 /aF7E}>S>WU_Z7E}Y>hz[1 mn:} }/NַCtQԧk#|Znc}n̷!榣 o ofU:1߆X>}9HzַكpQԧ#{Uc}G]߇>7E} }䓷:a֗ka(9#?}?Qb} b1Cn:sA OX+kwQ_rJ̻»8ow>.;;yZxg-V2PeV歲-0o N(DQˈFIꬼwNc4m2m GfS2o-em6~=||#3yqPayW UVYn\OOBeܯ% qk48S?ZvO@_v& TDǸNwpKٟ{Z珌g_/OE~Ol{ԱMaǸ /f{_؀4}]ɸ:ݛ!kK7{Z<: %N9I훕Njt|wwW*ErV)(sUOnWWӬkל>곙~r}~pS ]ܰI?} [e9cA|Wg3 ^[sWV[3愷K )tkcDWWr[?zvDd b  c $A? ?3"`?2F(;p\nHK`!F(;p\n@xcdd``fbb``baV d,FYzP1n:&B@?b 30|cꁪaM,,He`x7Ӂ`'0LYI9 sy$|bX|"+|-M[T T2i^yd? 6׶̬CiR jO)n0~ bF ,@{C2sSRsy #sNE --M0(2BƟ Ʒ5,HaqxdB%,A?`H--ʇr,CFQv>z0RxvK8!|N(ߍdk@>4p8Eʅ+KRsւlfPdk{> 1 3X?9ͩIt has been a long known fact that Croatians save in foreign currency, trying to avoid domestic currency as the currency of choice for savingsIKT Podgrupa Normal.dotIKT Podgrupa2Microsoft Office Word@F#@H@Hl՜.+,D՜.+,|8 hp DocumentSummaryInformation8|CompObjs Privredna banka Zagreb d.d.A It has been a long known fact that Croatians save in foreign currency, trying to avoid domestic currency as the currency of choice for savingsIt has been a long known fact that Croatians save in foreign currency, trying to avoid domestic currency as the currency of choice for savings NaslovTitle 8@ _PID_HLINKSA<MQ)http://ideas.repec.org/s/bla/restud.htmljwN<http://ideas.repec.org/a/bla/restud/v73y2006i3p793-821.htmlTK)http://ideas.repec.org/s/eee/moneco.htmlRFH>http://ideas.repec.org/a/eee/moneco/v53y2006i8p1761-1800.htmlPE)http://ideas.repec.org/s/aea/aecrev.html!bB;http://ideas.repec.org/a/aea/aecrev/v63y1973i3p326-34.htmlS?)http://ideas.repec.org/s/eee/jetheo.htmlb:<;http://ideas.repec.org/a/eee/jetheo/v4y1972i2p103-124.htmlfmailto:neven.vidakovic@pbz.hrfmailto:neven.vidakovic@pbz.hr      !"#$%&'@@@ NormalCJ_HaJmHsHtH>A@> Zadani font odlomkaVi@V Obi na tablica4 l4a .k. Bez popisa>@> 2 G Tekst fusnoteCJaJ>&@> 2 GReferenca fusnoteH*6U@6 s%g Hiperveza >*B*phBE@gEIG5     GG  "& "& "& "& "& "& "& "& "& "& "& "& "& "& "&|B&9/l6:DKcVd\f6p ~6xu   h {CDLMAMtuvwxyz{|A ( BxHIFzC!"%&(6)7)8)M)N)*****....9/:/R/T/1f1~112222R6S6k6l677777787Y7Z7[777788*8+8F8G8_8`8a88888:h;i;j;;;= ?z@%ACFFFFHIHINIfIgI^J_JwJxJKKKKKOQQQQT,WXXXX3Z[[[[<\=\?\d\e\r]]]]]R_`b"dffffKh2jlmmmmmnn+ooo6ppp4qGrsstuv|w xQxyyyz{l|H}} ~~;sȀ)%%%% %%%%%%%%%%%%%%%%%%%%%%%%%%%"%%%%"%,%%"%{ %%{ %%O %%{ %"%%"%"% %{ %O %{ %%%%%O %%%K%%q$%%%%{ %%%%"%%K%%{ %%7P%%E(%%?%%%%,%%%%%%%w6%%%%%%%%% %%%%%y %%%%%%%%%O %O %% %%%%%O %{ %%%%%{ %%Ӂ%%%{ %%w6%%,%%%%%% %O %%?%%"%O %%%%%%~&%%%{ %%%%%O %O %O %%%%%%"%%%{ %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%v:%v:%v:%v:%v:CDLMAMtuvwxyz{|A ( BxHIFzC!"%&(6)7)8)M)N)*****....9/:/R/T/1f1~112222R6S6k6l677777787Y7Z7[777788*8+8F8G8_8`8a88888:h;i;j;;;= ?z@%ACFFFFHIHINIfIgI^J_JwJxJKKKKKOQQQQT,WXXXX3Z[[[[<\=\?\d\e\r]]]]]R_`b"dffffKh2jlmmmmmnn+ooo6ppp4qGrsstuv|w xQxyyyz{l|H}} ~~;sȀ)0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@00@00000000@00@0@0@000@0@000@0@0@000@0@0@0@00@0@0 0@0@0@0 0@0@0@0 0@0@0@0 0@00@0 0@0@0@0 0@00@0 0@0@0@0@0@0@0@0@0@0@0@0 0@0@0@0 0@0@0@0@0 0 @0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0 0 @00@0 0 @00@0@0 0 @0@0@0@0@0@0@0@0@0@0 0 @0@0@0@0@0@0@0@0@0@0@0@0@0@0 0@0@00@0@00@00@00000@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@000@0@0@0@0000h000h000h000h000CDLMAMtuvwxyz{|A ( BxHIFzC!"%&(6)7)8)M)N)*****....9/:/R/T/1f1~112222R6S6k6l677777787Y7Z7[777788*8+8F8G8_8`8a88888:h;i;j;;;= ?z@%ACFFFFHIHINIfIgI^J_JwJxJKKKKKOQQQQT,WXXXX3Z[[[[<\=\?\d\e\r]]]]]R_`b"dffKh2jlmmmmmnn+ooo6pp4qGrsstuv|w xQxyyyz{l|~@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@ 0@0@0@0@ 0@0@0@0@ 0@0@0@0@ 0@0@0@0@ 0@0@0@0@ 0@0@0@0@ 0@0@0@0@0@0@0@0@0@0@0@0@ 0@0@0@0@ 0@0@0@0@0@ 0 @0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@ 0 @0@0@0@ 0 @0@0@0@0@ 0 @0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0h00@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@0@ 0@ 0j0/0@0 >F !#'06w= @DIaNSTU]fi'sHz؋8NFIKLMNOPQRSTUVXYZ\]_`acdeghAQUzifGJW[^bfHm(Ul***...:/N/P/f1z1|1222S6g6i6777A7U7W7g7{7}77778&8(8G8[8]8888NIbIdI_JsJuJKKKXXX]]]Jqqqqqr`rrrrsUssHt^tatttuKuauduuuXX::::::::::::::::::XXXXXXXXA( A8 A}AtAA4AA}AAt~A¹oA\A J K Q l u v { | }  JMNPQTU]^acfghktu|}z38z}~^hEL!"*+.4<=FIRS\GRlw|   & 7 B C I J L T \ [$b$x%%&&&&&&&&&&&&&&&'''''''''''''''''''''' ((((( 999!95999u>|>UUUUUU0]3] ^ ^ff\hdhvi|i}iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiijjjjjjjjjjjjjkk kkkkkkkllmmmmnn%n*n,n0nnnnn5o=oooooopp pp!p"p*p+p5pppppAqCqrrVrXrttuuuvvvvvvvvvvqw{w|wwwwwwwwwwwwwwwwwwwwwwwwxx!x*x+x3x4x6x=xCxDxIxZx_xhxnxoxwxxx~xxxxxxxxxxxxxxxxxxxxxxxxxxx yyyy y'y(y0y1y3y4y@yAyIyLyUykytyuy~yyyyyyyyyyyyyyyyyyyzz {{j{q{s{{{{{U|e|l|s|0}7}9}A}H}O}}}}}~~\eflquQ[*.Ml@A ' ( * x BHp!$y- N Q v y !!""T"V"&'''(5)(,-,-- .$.x6{6oL}LQQ ^^z^}^-`/`bbccfJhKhhii3lllmmnnnooWoZoooIqJqrrsssstttcuuuvvvvvvdwqw|wwww=xCxyy y"ydyjyyy{{ ~Q~{~~~~~~~Ve::::::::::::::::::::::::::::::::::::::::::::::::::::::::::DMA|3 [ ( Bz- c (8)B1f167+8G8a88KKDggKhhlmmst/9Xc D'f ::v*>h ^`hH.h ^`hH.h pLp^p`LhH.h @ @ ^@ `hH.h ^`hH.h L^`LhH.h ^`hH.h ^`hH.h PLP^P`LhH.h ^`hH.h^`5\hH. pLp^p`LhH. @ @ ^@ `hH. ^`hH. L^`LhH. ^`hH. ^`hH. PLP^P`LhH.:: D' .B\wE% r/> A M=WUbe3*.Vweui =%F%Om%*&Q&}+/,P-4e-y-V/09L2?5.O8@Q@cb@xA@BJCU D,EG2 G\+GpkG"H#I+IJML)bLXM9O}P2=QtR}RkHWBK\c](O` aiIahba^cIQfs%gCjHpCq!stt1t?vLRwKyQyezX}_b}T~}0~@DfleZ ]"\ EyPW`SdG#a#d(@,2yn{ .+W2 $R"Q&8=2) xyY+?u%~ _T_3FhVhN.4!R"ys$ %ADZ4R yk}o R RlyKlBoz66ydlv'F9%wfs~@jj$jj UUUUUUU U(U)o*o+o5o7FGIJLNOSTVWYZf@  D@$L@.`@46p@@@D@P@d@hj@p@v@z@~@UnknownGz Times New Roman5Symbol3& z Arial"qF&F&lAlA!nn>4d 3QHX)?2It has been a long known fact that Croatians save in foreign currency, trying to avoid domestic currency as the currency of choice for savings IKT Podgrupa IKT Podgrupa    F!Microsoft Office Wordov dokument MSWordDocWord.Document.89q